Flame-Retardant and Fire-Sensing Packaging Papers Enabled by Diffusion-Driven Self-Assembly of Graphene Oxide and Branched Polyethyleneimine Coatings

Abstract

Paper has gained popularity as a packaging material due to its reduced environmental impact compared with non-degradable alternatives. However, its flammability poses safety risks, prompting research on enhancing its flame retardancy. This work introduces a diffusion-driven self-assembly strategy (DDSAS) to create a functional graphene oxide (GO) coating on various packaging papers. DDSAS involves infiltrating the paper’s cellulose microfiber network with branched polyethyleneimine (b-PEI), which binds firmly to cellulose microfibers. Electrostatic interactions between GO and b-PEI then drive GO assembly into a densely stacked, layered structure on the paper surface. This GO structure provides a physical barrier against flames and generates incombustible gases (CO₂, H₂O, NO₂, and NO) when heated, diluting the surrounding oxygen concentration and acting as a heat insulation layer. These factors increase the flame retardancy of treated papers ten-fold. Additionally, the gradual reduction of GO upon heating forms reduced graphene oxide (rGO) on the paper, significantly increasing its electrical conductivity. As a result, the flame-retardant papers not only prevent the fire from spreading but can also act as fire sensors by triggering an alarm signal at the early stages of contact with fire. In summary, this work offers a rational strategy for designing and manufacturing flame-retardant paper packaging materials.

1. Introduction

Environmental and social sustainability requirements have promoted the development of greener materials that are more environmentally friendly and have lower carbon footprints. Packaging materials face these development requirements. The most widely used packaging materials on the market today are glass, plastic, metal and paper. In particular, plastics are widely used in packaging for their light weights, good mechanical strengths, and good air tightness. However, plastics are difficult to recycle and do not degrade quickly [1,2]. Therefore, most plastic packaging ends up in landfills [3], leading to the accumulation of microplastics in the soil and ocean. These microplastics remain in the natural environment for a long time and accumulate in organisms and human bodies, thereby directly affecting the ecological environment as well as the health of humans and other living organisms [4,5,6]. Therefore, in recent years, the use of paper as a packaging material has gradually increased. The main reason for the growing proportion of paper packaging is that as a material with natural origins, paper has significant advantages over other packaging materials in terms of recyclability and biodegradability [7,8,9,10]. However, paper materials tend to be highly flammable and need to be treated so that they are flame retardant. Researchers have developed a series of methods and techniques to make paper more flame retardant. The most widely adopted strategy is to add flame-retardant chemicals to the paper, such as phosphoric acid [11], phytic acid [12], inorganic materials [13,14], etc. However, the introduction of these flame retardants usually worsens the mechanical properties of the treated paper [15]. Moreover, in some packaging materials designed for food applications, the type and quantity of flame-retardant substances that can be used are limited. Therefore, modification of paper using a coating process is a preferred alternative. Such modification processes not only preserve the inherent mechanical properties of the paper, but also provide a physical barrier to prevent the outer coating material from seeping through the paper and coming into contact with the packaged item. Therefore, a series of coating materials (such as PVC [16], PVA [17] and PLA [18]) have been used to develop flame-retardant packaging materials and are also used in foods, medicines, and other fields. However, these materials still have many shortcomings. For instance, polymeric materials release a large amount of toxic gas when heated, which can directly pollute the environment or endanger the safety of packaged items. The generation of toxic gases can also lead to a secondary disaster. In addition, packaging materials are usually used in warehouses or various modes of transportation since they are typically used to encase bulk commodities. Therefore, in addition to providing structural support, packaging materials that are smart have the added advantage that they can provide early warnings of a potential fire emergency.

In this respect, GO-based coatings have many advantages. GO is rich in hydrophilic functional groups, allowing it to be dispersed into aqueous solutions and further processed into two- or three-dimensional architectures [19,20]. GO also is biocompatible and has antibacterial properties [21,22,23,24]. Moreover, as a carbon-based material, GO will not burn spontaneously or generate byproducts that can have negative effects on the environment. More importantly, in the process of heating, GO can be reduced or partially reduced to rGO and transition from being an insulator to a semiconductor [25,26]. This chemical transition provides a means of creating an early warning signal for fire. Based on these properties, the present work develops a diffusion-driven self-assembly strategy (DDSAS) to functionalize packaging papers using GO and b-PEI as starting materials. The DDSAS starts by incorporating the branched polyethyleneimine (b-PEI) into the cellulose microfiber network of the paper. Then, the electrostatic interactions between GO and b-PEI drive GO assembly and the formation of a densely stacked and layered coating on the paper surface. In this coating, GO serves as the skeleton and the PEI polymer acts as the “adhesive”. The prepared ultra-thin coatings not only significantly enhance the flame-retardancy of the paper, but can also provide an alert to a fire emergency. Heating the coating results in the reduction of GO to form reduced-GO (rGO), and the coating changes from an insulator to a semiconductor as a result of this transformation.

2. Materials and Methods

2.1. Materials

Graphite powder (SP-1, Bay Carbon, Inc., Bay City, MI, USA). Polyethyleneimine (50 wt.%) and KMnO₄ (potassium permanganate, AR) were supplied by Shanghai Macklin Biochemical Technology Co., Ltd. (Shanghai, China). N, N-Dimethylformamide (DMF, AR), H₂SO₄ (sulfuric acid, 95%), H₂O₂ (hydrogen peroxide, 30%) and HCl (hydrochloric acid, AR) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). K₂S₂O₈ (potassium persulfate, AR) and P₂O₅ (phosphorus pentoxide, AR) were provided by Aladdin reagent (Shanghai, China) Co., Ltd.). Duplicating paper, A4; kraft paper, A4, 120 g; cardboard from the packaging carton of Whatman dustless paper; and filter paper (Whatman, grade no.3, Maidstone, UK) were also used. All the materials and reagents were used as received without further treatment.

2.2. Methods

2.2.1. Synthesis of Graphene Oxide

Graphene oxide was synthesized using a modified Hummers method [27], consisting of three main steps: pre-oxidation, oxidation, and exfoliation. In pre-oxidation, concentrated H₂SO₄ (20 mL) was added to a 500 mL round-bottom flask in an oil bath at 80 °C. K₂S₂O₈ (3.75 g) and P₂O₅ (3.75 g) were then added, and the mixture was stirred until fully dissolved. Graphite powder (10 g) was subsequently introduced, and the solution was maintained at 80 °C for 4.5 h. The mixture was cooled, diluted, and vacuum filtered. The collected product was rinsed with DI water until the filtrate pH was neutral and dried in a vacuum oven at 30 °C. During oxidation, concentrated H₂SO₄ (187.5 mL) was added to a 1 L Erlenmeyer flask and cooled to 5 °C. Pre-oxidized graphite powder and KMnO₄ (25 g) were added, and then the flask was transferred to a 35 °C water bath for 3 h before cooling. The suspension was carefully diluted with DI water and H₂O₂ was added dropwise, yielding an orange precipitate of oxidized graphite. The product was further treated with HCl (3%) and centrifuged at 8850× g for 10 min. This process was repeated three times before the precipitate was redispersed in HCl (3%) to a total volume of 500 mL. The final exfoliation step involved removing HCl from the oxidized graphite suspension by centrifugation and exfoliating it into GO nanosheets via sonication in water at a concentration of 1.5 mg/mL. The exfoliated GO nanosheets were concentrated by centrifugation at 17,700× g for 2 h. The concentration of the GO suspension was determined by freeze-drying a 1 mL sample.

2.2.2. Fabrication of GO/b-PEI Composite Coatings on Substrates

The fabrication of the GO/b-PEI composite was modified based on a previously reported diffusion-driven layer-by-layer assembly [28]. Firstly, a 4 mg/mL GO suspension was prepared in a DMF: water (1:1) solution, denoted as GO dispersion. Next, an A4 copy paper was placed on the surface of a 10% (w/w) b-PEI aqueous solution in a rectangular container. After 5 min, the copy paper was removed and dried in an oven at 60 °C for 20 min. The b-PEI-coated copy paper was then placed on the surface of the GO suspension for 5 min to create a GO/b-PEI composite coating. Due to surface tension and the paper’s light weight, the paper does not submerge in the b-PEI solution and GO suspension within the experimental timeframe, resulting in coating formation solely on the surface in direct contact with the GO suspension. The coated copy paper was rinsed gently with deionized water to remove any residues and dried in a 60 °C oven for 1 h, forming a dense GO/b-PEI composite coating with a layer-by-layer structure. This diffusion-driven process can be applied to other surfaces, such as kraft paper, cardboard, or silk fabric, that can be coated with b-PEI. The fabrication of a GO/b-PEI composite coating on silk fabric follows the same steps as for copy paper. Due to the silk fabric’s loose structure, the GO/b-PEI composite coating can be partially peeled off to obtain small fragments measuring 0.5 to 3 cm in length.

2.3. Characterization Methods

Scanning electron microscopy (SEM) images of the GO/b-PEI composite coatings on duplicating paper were detected using a JSM-7800 Prime system, while SEM images of the paper were taken with a JSM IT-500HR/LA (JEOL, Tokyo, Japan). Insulated samples were coated with Au nanoparticles to enhance conductivity for SEM characterization. TEM images of the GO nanosheets were obtained using a JEM-1400-plus (JEOL, Tokyo, Japan), with the GO suspension diluted in ethanol (0.01 mg/mL). After applying a drop of the diluted suspension to a TEM grid, it was dried at 30 °C for 30 min. XRD spectra were obtained using a Bruker D8 diffractometer (8° to 60°, Cu-Kα radiation, 40 mA, 40 kV, Bruker, Billerica, MA, USA). GO powder for XRD was prepared by freeze-drying and pressing into a film. TGA-MS spectra were acquired with a Netzsch STA449F3@QMS 403 (Bayern, Germany) in air (flow rate 30 mL/min). A 4.31 mg GO/b-PEI film was heated from 30 °C to 1000 °C at a rate of 10 °C/min. GO/b-PEI films for FT-IR and TGA-MS characterization were obtained by peeling them off silk fabric surfaces. The resistance–time curve of a 2 cm × 2 cm piece of cardboard with GO/b-PEI composite coating was collected using a Keithley DMM-6500 digital multimeter (2-wire mode, 0.1 s resolution, Cleveland, OH, USA). The resistance of the GO/b-PEI composite coating on duplicating paper after heating was tested using a Keithley digital multimeter with a probe distance of 1 cm. An alcohol burner and lighter were used as the flame source in the flame-retardant property testing experiments and the fire alarm testing experiment.

3. Results

3.1. Design and Fabrication of GO/b-PEI Composite for Coating on Duplicating Paper

Scheme 1 shows the experimental protocol for the in situ self-assembly of the GO/b-PEI composite coating on a paper surface using DDSAS.

First, the paper substrate was placed in a b-PEI solution. Driven by osmotic pressure, the b-PEI molecules (a positively-charged polyelectrolyte) entered into the cellulose microfiber network (negatively charged) of the paper. After approximately 5 min, the paper was removed from the b-PEI solution and dried. The b-PEI molecules were retained in the paper after this process, and the electrostatic interactions between the paper microfibers and b-PEI resulted in strong adhesion of the two components. Subsequently, the b-PEI-immersed paper was placed in a GO dispersion. The electrostatic interactions between the GO and b-PEI embedded in the paper form a stable complex at the solid–liquid interface. The GO/b-PEI composite formed at the interface is semipermeable and allows continuous diffusion of the b-PEI molecules across the interface, which leads to gradual development of a foam-alike macrostructure [28]. This self-assembly process can occur in any configuration as long as a stable liquid–liquid interface is formed. Thus, the process is highly versatile and can be tailored to the final shape of the composite. For example, large area films of the porous GO-based scaffolds can be easily prepared by inducing assembly on a large substrate. The final GO/b-PEI-coated (GPC) papers were obtained by washing and drying the resultant papers.

3.2. Characterization of GO/b-PEI Composite Coating

As demonstrated in Figure 1a, GPC-paper with dimensions of 20 cm × 21 cm was fabricated using the self-assembly strategy presented here. The composite coating was highly uniform, with strong interfacial adhesion in the obtained GPC-paper. Furthermore, the electrostatic interactions between the positively charged b-PEI and negatively charged cellulose resulted in good adhesion to the substrate, enabling the formation of GO/b-PEI coatings on various paper materials, such as copy paper, kraft paper and filter paper. Figure 1b displays the characteristic absorption peaks of GO at 3370 cm⁻¹, 1737 cm⁻¹, 1228 cm⁻¹, and 1057 cm⁻¹, which are attributed to –OH, C = O (carboxylic acid), C–O (alkoxy), and C–O–C (epoxy) groups, respectively [29,30]. Additionally, the peak at 1625 cm⁻¹ is an overlapped band due to bending vibration of –OH, intercalated water, and stretching vibration of C=C. The characteristic absorption peaks of b-PEI at 2945 cm⁻¹ and 2841 cm⁻¹ are assigned to the symmetric and antisymmetric stretching vibrations of the –CH₂– groups. The peaks at 3296 cm⁻¹, 1642 cm⁻¹, and 1463 cm⁻¹ correspond to N–H stretching, N–H bending, and C–H groups, respectively [31,32]. In the GO/b-PEI composite, the absorption peaks of –CH₂– groups shifted to 2919 cm⁻¹ and 2815 cm⁻¹, and the peaks of -NH₂ and -NH groups disappeared. The wide peak due to –OH at 3370 cm⁻¹ and the peak at 1737 cm⁻¹ belonging to C=O of the GO sheets also disappeared [33,34]. It is speculated that the strong electrostatic interactions between the functional groups of GO and b-PEI caused the shifting and disappearance of their corresponding absorption peaks in FT-IR spectra.

Figure 2 further compares the differences in the morphology and microstructure of GPC-paper and copy paper before DDSAS processing. The color of GPC-paper varied from white to light brown, indicating that GO and b-PEI were successfully assembled on the surface of the copy paper. In addition, the GO/b-PEI coating did not peel off the paper after rubbing or immersing it in water, which showed the strong interfacial adhesion between the GO/b-PEI coating and copy paper. The good interfacial adhesion was further confirmed by scanning electron microscopy (SEM). As shown in Figure 2a, there were obvious gaps and voids on the surface of the untreated copy paper. The cellulose microfibers were stacked and formed layers, and there were air gaps between microfiber layers. In contrast, after DDSAS processing, the gaps and voids between the cellulose microfibers were filled by GO/b-PEI. As a consequence, the GPC-paper was denser, and no obvious structural defects were observed in the SEM images in Figure 2b. The adhesion between cellulose microfibers was also enhanced. Cross-sectional SEM images showed that the GO/b-PEI coating had a thickness of about 750 nm and was tightly attached to the copy paper as there were no obvious voids between two layers (Figure 2c,d). The X-ray diffraction (XRD) patterns (Figure 2e) measured for the GPC-paper revealed that the characteristic peak corresponding to the layer repeat distance in GO at 10.9° disappeared, indicating that the GO had been well dispersed in the b-PEI. In addition, no other characteristic peaks were seen in the XRD pattern of the GPC-paper at scattering angles ranging from 11° to 60°, which indicated that the graphite had been exfoliated into a single layer or few layers. In addition, the coating layer was found to have minimal impact on the breathability properties of the original paper (Figures S1 and S2) while also increasing the water-resistant ability of one side of the packaging material (Figure S3), thereby enhancing its stain resistance. The GPC-paper showed good resistance to water immersion and ultrasonic treatment (Figures S4 and S5). The wash water color and clarity remained unchanged, and only a minimal coating detachment (below 2.02%) was observed in the GPC-kraft paper and GPC-cardboard (Table S1 and S2, see Supplementary Materials) after ultrasonic treatment.

3.3. Flame-Retardant Property of Paper Modified by GO/b-PEI Composite Coating

To verify that the GO/b-PEI coating enhanced the flame retardancy of the paper, we carried out a series of comparative combustion tests. It can be seen that GPC-copy paper did not burn even upon exposure to a continuous flame for 1 min. In contrast, the untreated copy paper ignited at 6 s and had completely burned to ashes at 16 s (Figure 3a,b). Moreover, careful observation of the GO/b-PEI coating on the GPC-copy paper revealed that the color of the coating changed from brown to black and then to silver gray upon continuous exposure to the flame. The color changes also corresponded to changes in the resistance of the GO/b-PEI coating from 60 MΩ (brown) to 30 MΩ (black) and a further decrease to 50 kΩ (silver gray) in Figure 3c,d. This result indicated that during the combustion process, GO was gradually reduced to rGO at high temperatures. SEM characterization revealed that the microstructure of the GO/b-PEI coating was well preserved even though the macroscopic color of the coating varied from brown to black. While the SEM images suggested that the coating was still intact when it turned silver gray, it was obviously wrinkled, indicating that some of the b-PEI and cellulose had decomposed at the higher temperatures (Figure 3c). In addition, some bulging structures were clearly observed in the SEM images (Figure 3d). It is speculated that these bulges resulted from the gas generated during the heating process. The generated gas was not fully released due to the dense packing of the GO layers and thus formed a bubble between the GO coating and the cellulose on the paper surface. Therefore, the coating was not as tightly bound to the cellulose substrate after heating.

The hypothesis was also supported by the cross-sectional SEM images. As shown in Figure 4a, the decomposition of b-PEI and reduction of GO in GPC also led to a decrease in the thicknesses of the coating and paper, and as a result, the cellulose substrate was not well resolved in the cross-sectional images of the coating post-combustion. The thickness of the silver-gray coating had decreased to ~150 nm. However, the rGO remained tightly stacked, and the cross-sections of GO changed from being relatively straight to a wavy. The structural evolution of the GO/b-PEI composite was also confirmed by XRD and FT-IR spectra. As shown in Figure 4b, the characteristic peak of GO shifted from 10.9° to 21.5° and 23.8°, where the higher angles were closer to those seen in graphite (26°). In the FT-IR spectra (Figure 4c) of GO/b-PEI heated and turned to silver gray, the peaks at 2919 cm⁻¹ and 2815 cm⁻¹ of –CH₂– groups and at 1281 cm⁻¹ of C–N groups from the polymer chains had disappeared, indicating the decomposition of b-PEI. The remained peak at 1548 cm⁻¹ corresponded to the stretching vibration of C=C, which came from the restoration of benzyl rings on reduced GO sheets. The peak of C–O–C (ethers) stretching vibration shifted from 1057 cm⁻¹ to 1148 cm⁻¹, which implies structural changes in the GO sheets after heating [20,30,35].

In order to verify the universality of the flame-retardant performance of the GO/b-PEI coating, similar flame tests were performed on kraft paper coated with GO/b-PEI. In these tests, the GPC-kraft paper was first folded into a packaging box and then continuously heated with a flame. As shown in Figure 5, the untreated kraft paper ignited and burned within 7 s, while there was no obvious ignition after 100 s in the region over the box covered with the GO/b-PEI coating.

3.4. Flame-Retardant and Alarm Mechanisms of GO/b-PEI Composite Coating

TGA-MS spectra were used to further reveal the flame retardancy mechanism of the GO/b-PEI coating (Figure 6). The TGA curves showed that the main decomposition process in the GPC-paper occurred at around 800 °C. The corresponding MS analysis revealed that the main gases were generated during the combustion process in the temperature range of 30–800 °C were H₂O, CO₂, NO₂ and NO. These gases are non-combustible, and therefore, diluted the oxygen concentration or even isolated oxygen in the atmosphere near the paper, thereby inhibiting the rapid spread of the flame. In addition, some of these gases were trapped at the interface between the paper and GO and formed a non-combustible gas layer or air layer that was evidenced by the bulging structures in the GO/b-PEI coating. These gas pockets acted as an insulating layer and reduced the temperature of the paper, which in turn, significantly prolonged the ignition time. Moreover, the formation of the bulged structure also prevented the paper from coming into direct contact with the flame.

More details regarding the combustion process and corresponding flame retardancy mechanism were revealed through a comparison of the TGA and MS results. Specifically, below 130 °C, the mass loss was due to the release of free water or bound water in GPC-paper. Between 130 and 200 °C, the MS results showed both H₂O and CO₂ were present, indicating the mass loss was due to decomposition of the oxygen-containing functional groups in GO and the evaporation of free water. As the temperature increased above 200 °C, the water bound to b-PEI was lost and the b-PEI started to decompose, thereby producing gases such as CO₂, H₂O, NO and NO₂. GO was also gradually reduced to rGO in this temperature range. The reduction of GO also produced CO₂, which is why large amounts of CO₂ were observed even at temperatures as high as 600 °C.

Based on the above observations, the flame alarm mechanism of the GO/b-PEI composite coating is shown in Scheme 2. With the action of the flame, the non-conductive GO/b-PEI composite is thermally dehydrated at first and then carbonized. During the carbonization process, the decomposition of b-PEI polymer chains and the reduction of GO nanosheets proceed simultaneously, leading to the massive release of non-combustible gases (CO₂, NO and NO₂) that fill the space between the layers of reduced graphene oxide. As the gases diffuse into the air and the GO nanosheets restore the aromatic ring structure, the distance between the rGO layers reduces further, resulting in the formation of a densely packed layer-by-layer structure of rGO thin film with excellent conductivity.

3.5. Early Fire Alarm Application of GO/b-PEI Composite Coating

As previously mentioned, the electrical resistance of the GO/b-PEI coating layer changes when heated. We assessed the detailed resistance changes produced by the GPC-paper after exposing it to an open flame (Figure 7a). In particular, as the coating changed from black to gray, the resistance dropped to 13 MΩ. The coating created a flame-retardant layer with semiconducting properties, in which the resistance tends to decrease as the temperature rises. Consequently, when the GPC-paper was exposed to a flame, its measured resistance experienced a sharp decrease, followed by a gradual rise after the flame was removed. This significant decline in electrical resistance could act as an early warning “signal” for fire detection. When the GPC-paper was exposed to flame again, the resistance decreased from 120 MΩ to 3.6 kΩ after 15 s of continuous heating.

We built a complete test device to demonstrate the early warning capabilities of GPC-paper based on the sharp decrease in resistance that corresponds to early flame phenomenon (Figure 7b). When the GO/b-PEI-coated cellulose paper was exposed to flame, the electrical resistance of the paper was detected and fed back to the signal processing platform. Once the resistance change matched a preset threshold (i.e., 1000 Ω in this work), the alarm was automatically triggered. Additionally, Wi-Fi and internet application protocols could be involved in the fire alarm platform (Figure 7c), enabling the fire alarm signal to be sent to various external ports, such as a cell phone or control center. In practical applications, when a small flame occurs and touches the GO/b-PEI-coated packaging, although the coating layer is too thin to act as a complete thermal insulator, the flame-retardant property prevents the fire from spreading and triggers the alarm signal. This would attract the administrator’s attention to come and check before the fire grows bigger.

4. Conclusions

In summary, GO/b-PEI coatings with good flame-retardant properties were prepared on various paper surfaces by DDSAS. The interactions between GO and b-PEI as well as the hierarchical structure of the resultant GPC-paper were analyzed through combination of FTIR, XRD and SEM. The results revealed that the DDSAS process resulted in an ordered assembly of GO into layered structures on different paper surfaces. Furthermore, it was also clear that the assembled GO/b-PEI coatings were tightly attached to the different types of paper. This structural advantage, together with the flame retardancy of GO and b-PEI, endowed the GPC-papers with good flame retardancy. For instance, the flame retardancy of copy paper and kraft paper increased by factors of 10 and 100, respectively. TGA-MS further revealed that the flame retardancy of the GO/b-PEI coating was due to the generation of incombustible gases upon heating that diluted the oxygen content in the environment immediately surrounding the paper. Moreover, some of the incombustible gases were trapped between the paper and coating due to dense stacking of the GO layers, and the trapped gases formed bubbles that acted as adiabatic air barriers to slow down the heat transfer rate to the paper. Importantly, the DDSAS exhibited versatility by being applicable to various packaging papers with distinct mechanical properties. It successfully preserved the mechanical performance of these papers while enhancing their functionality as both flame retardants and indicators for detecting fire emergencies. Additionally, we constructed a comprehensive testing apparatus to showcase the early warning capabilities of the GPC-paper. The resistance of the GPC-paper exhibited a sharp decrease corresponding to the occurrence of early flame phenomena. The preliminary studies presented in this paper highlight the advantages and potential of GPC-paper, serving as a foundation for its commercial development and practical application.