Introduction

Nowadays, due to the shortage of petroleum resources as well as the serious environmental issues caused by petroleum-based polymers, the development of bio-based polymer material has attracted extensive attention from both academia and industry (Jiang et al. 2018; Long et al. 2018). Cellulose is one kind of naturally available bio-based polymer on earth. It is a linear syndiotactic homopolymer composed of glucose residues that are linked together by β-(1 → 4)-glucosidic bonds (Pan et al. 2014; Suflet et al. 2006). Compared with petroleum-based polymers, cellulose possesses many superior properties such as low cost, low density, renewability, and biocompatibility (Habibi et al. 2010). As a common chemical raw material, cellulose’s industrial applications in paper, fabrics, and building materials fields can date back to thousands of years (Long et al. 2018). More and more cellulose derivatives have been developed and studied. Cellulose nanofiber (CNF) foams prepared by decomposition of wood fibers are natural polymer materials, which have attracted considerable attention due to their high specific surface area, non-toxic and biodegradable characteristics (Guo et al. 2018).

However, as a limitation, cellulose is flammable because of its inherent chemical structure. When it is ignited, cellulose and cellulose derivatives pyrolyze to produce many volatile products and flammable gases, which may cause some safety problems. Because halogen-containing flame retardants release some toxic and harmful substances during combustion, green, environmentally friendly, and halogen-free flame retardants are demanded. Han et al. (2015) synthesized flame retardant cellulose foams by in situ synthesis of magnesium hydroxide nanoparticles in cellulose gel nanostructures, which showed excellent flame retardancy and excellent insulation performance with the low thermal conductivity. Over the past decades, layered double hydroxides (LDHs) may be regarded as a promising new green flame retardant for polymer applications owing to their unique chemical composition and layered structure, which provide them excellent flame retardancy and smoke suppression properties (Gao et al. 2014). Many kinds of LDH used as flame retardant containing both phosphorus and metal elements were synthesized, such as biobased eugenol modified LDH (SIEPDP-LDH) (Li et al. 2015), the APTS grafted/tripolyphosphate intercalated ZnAl LDHs (Xu et al. 2016), MgAl-layered double hydroxide (MgAl-LDH) (Luo et al. 2020), and so on. The doping of metal ions into polysaccharide foams expands the application fields of cellulose foams. Hu et al. (2019a, b) prepared bulk Al-doped carboxymethyl cellulose foams with low density, perfect porosity outstanding weight-bearing capacity, and flame retardancy. Cheng et al. (2020) successfully prepared cellulose nanofiber (CNF)/zinc borate (ZB) foams, and the improved flame retardancy of CNF/ZB foams will not affect their original thermal insulation function. The addition of inorganic hydroxide and metal ions flame retardant could enhance the flame retardancy of cellulose foams. But physical blending nanocomposite methods and most of the fabrication processes in these studies were intricate and required plenty of inorganic fillers (often more than 40%) to achieve excellent flame retardant. Therefore, it is necessary to find better and feasible methods to prepare cellulose foam with high flame retardancy without sacrificing other properties.

The flame retardant containing phosphorus shows higher flame retardant efficiency. Nitrogenous or metallic compounds are also generally combined with phosphorus flame retardants for application, where P\N elements play a synergistic flame retardant role (Liu et al. 2016; Niu et al. 2020). Phosphoric acid can be utilized to prepare phosphorylated cellulose derivatives, which can satisfy the need for biocompatibility and degradability rather than only flame retardancy (Bauer et al. 2017). Du et al. (2020) synthesized two-dimensional layered black phosphorus (BP) nanosheets with excellent properties from black phosphorus (BP) and combined them with cellulose to receive the CNF/BP hybrid foams material. Furthermore, phosphazene was also proven as a typical flame retardant for cellulose. Yang et al. (2012) synthesized a durable flame retardant for cellulosic fabrics by using a phosphorus-based compound after a two-step process.

Phytic acid (PA), also known as inositol hexaphosphate, is rich in natural substances in nuts, cereals, beans, and other plants (Hu et al. 2019a, b). Phytic acid is also used to prepare multifunctional composited materials with many remarkable properties for some potential applications (Wang et al. 2021; Zhang et al. 2019). In our work, we chose phytic acid (PA) and guanazole (GZ) as biogenic materials and combined them with CNFs to obtain ideal flame retardant cellulose materials with low density, high porosity, and remarkable thermal insulating property. A series of composite foams with different CNF contents were obtained by the freeze-drying method. The microstructure, physical properties, hydrophobicity, thermal properties, and flame retardancy of the PA-GZ-CNF foams were investigated and discussed.

Materials and methods

Materials

CNF suspension (1.2 wt%) fabricated by TEMPO-mediated oxidation of pulp was purchased from Tianjin Haojia Cellulose Co. Ltd. (China). The length of CNFs was 0.8 ~ 2.0 μm, and the diameter was 5 ~ 20 nm. Phytic acid was bought from Sinopharm Chemical Reagent Co., Ltd. (China). Guanazole was offered by Bide Pharmatech Ltd. (China). Sylgard™ 184 silicone elastomer was supplied by the Dow Chemical Company (America). All the chemical reagents were used without any purification.

Synthesis of PA-GZ-CNF composite foams

A series of PA-GZ-CNF composite foams with CNF mass percentages of 100%, 85%, 80%, and 75% were prepared, respectively, where the mass ratio of phytic acid and guanazole was 1:1 in each sample. For example, to fabricate the PA-GZ-CNF-3 composite foam, 0.2 g of PA and 0.2 g of GZ were mixed with 5 mL of deionized water. Then, the solution was added to 100 g of CNF suspension and vigorously stirred at 60–70 °C for 1 h. After pouring into the mold, the mixed hydrogel was frozen at − 20 °C for 24 h and then freeze-dried under a vacuum for another 72 h to fully evaporate the water in the samples. All the samples were prepared by a similar process just with the different contents of PA and GZ. The formulations of CNF composite foams are listed in Table 1.

Table 1 Formulations of CNF and PA-GZ-CNF composite foams
Full size table

Hydrophobic treatment of CNF composite foams

The PA-GZ-CNF-3 composite foam was chosen to do the hydrophobic treatment. 6 g of Sylgard™ 184 silicone elastomer (part A) and 0.6 g of curing agent (part B) were dissolved in 300 mL trichloromethane. After soaking the foam in the solution for 5 min, the foam was placed on a clean glass plate and dried in the oven at 80 °C for 2 h, which was called PDMS-PA-GZ-CNF-3. The mass percentage increase was about 127% before and after hydrophobic treatment.

Characterization

The density (\(\rho\)) of composite foams was calculated by the ratio of mass to volume. The porosity was calculated according to formula (Medina et al. 2019; Yuan et al. 2017): porosity (%) = (1-\(\frac{\rho }{{\rho^{*} }}\)) × 100, \(\rho^{*} = 1.46\omega + 2.42 \times \frac{1 - \omega }{2} + 1.686 \times \frac{1 - \omega }{2}\). ω is the weight percentage of CNF in the foams. 1.46, 2.42 and 1.686 g cm−3 were the densities of cellulose, phytic acid and guanazole, respectively, for calculation in this formula.

The microstructure and elemental mappings of CNF composite foams were obtained by using an EVO MA15 scanning electron microscopy (SEM) These samples were previously sputtered with a layer of gold to increase conductivity before observation.

The LOI test was carried out on an HC-2 instrument (Jiangning, China) following the ASTM D2863-97. UL-94 vertical burning test was conducted on a CFZ-2-type instrument (Jiangning, China) according to the GB/T 8333–2008. The length and diameter of foams were 100 mm and 7 mm used in the UL-94 test.

The thermal conductivity of samples was evaluated on a thermal conductivity tester with a TC3000E device model (Xiaxi electronic technology co. LTD, Xi’an). All the samples were tested three times and the mean values were reported in this work.

Compression measurement was conducted on a CMT6104 Electromechanical Universal Testing Machine (MTS Systems Co., China) according to the ASTM D3575-14 standard. The sample used for compression measurement was a cylinder with a radius of 10 mm and a height of 40 mm. Three samples were repeated for each CNF composite foam for compression test.

The thermal degradation properties of CNF samples were researched by a TAQ5000 thermal analyzer instrument (USA). The samples were heated under nitrogen atmosphere from room temperature to 800 °C at a heating rate of 20 °C/min.

The combustion behavior of the samples was evaluated on a cone calorimeter (Su Zhou, China) according to the ISO5660-1. The square samples with the size of 100 mm × 100 mm × 3 mm were wrapped with aluminum foil and horizontally placed under a heat flux of 35 kW m−2.

The microstructure of the char residues was observed by Laser Confocal Roman Microscope (LabRAM HR Evolution, France) with a 514.5 nm argon-ion laser.

Thermo ESCALAB 250Xi X-ray photoelectron spectroscopy (XPS) was used to obtain the elemental composition of the char residue, which uses monochromatic Al Ka (hv = 1486.6 eV) as an X-ray excitation source.

The thermogravimetric analysis/infrared spectrometry (TG-FTIR) technique was applied to detect the evolution of gaseous products. The samples were performed on a Q5000 IR thermogravimetric analyzer interfaced with the Nicolet 6700 FTIR spectrophotometer. About 5.0 mg of the sample was heated from room temperature to 800 °C with a heating rate of 20 °C/min under a nitrogen atmosphere.

The contact angle experiment was performed on an SL200KB contact angle meter (USA KINO Industry Co., Ltd). The contact angle was measured by the sessile drop method and calculated by Young–Laplace fitting equation.

Results and discussion

SEM images revealed the surface morphology of the CNF composite foams. In Fig. 1, CNF foams after freeze-drying had more pores on the surface with higher porosity, and the size distribution of cellulose foam pore is not uniform. PA-GZ-CNF foams also exhibited a highly hierarchical porous structure, but the pores on the surface were less than pure CNF foam and the surface was slightly flat and compact. With the increase of PA-GZ concentration, the porosity of all the composite foams decreased slightly, which was still higher than 97% in spite of the increased density. In Fig. 2, the energy dispersive spectroscopy (EDS) elemental mapping was used to analyze the distribution of P and N elements in the selected region on the surface. PA and GZ were homogeneously distributed in the CNF foam, which did not affect the spatial structure of cellulose. The uniform distribution of PA and GZ was beneficial to improving the physical properties of the cellulose-based foams.

Fig. 1
figure 1

SEM images of a CNF b PA-GZ-CNF-1 c PA-GZ-CNF-2 and d PA-GZ-CNF-3 composite foams with different magnifications

Full size image
Fig. 2
figure 2

a EDS hierarchical image and b total surface spectrum of PA-GZ-CNF-3 foam, Elemental mappings of c nitrogen and d phosphorus in PA-GZ-CNF-3 foam

Full size image

Physical properties including density, porosity, and thermal conductivity are listed in Table 2. The pure CNF foam had a low density of 0.022 g/cm3 and displayed a high porosity of 98.53%. The density increased from 0.022 to 0.045 g/cm3 and the porosity reduced slightly from 98.53% to 97.07%, when the PA-GZ content increased from 0 to 25%. The thermal conductivity coefficient of the CNF composite foams was measured to determine their thermal insulating property. The addition of PA and GZ decreased the thermal conductivity of CNF composite foams from 0.0318 to 0.0298 W/m*K, indicating that the composite foams possessed remarkable thermal insulating properties. The mechanical properties of CNF and PA-GZ-CNF composite foams were evaluated by compression measurements. With the increase of PA and GZ content, the compressive strength of PA-GZ-CNF composite foams increased gradually. The incorporation of PA and GZ played an important role in the improvement of mechanical properties because PA and GZ could occupy some pores to prevent the porous structure from destruction. However, the resilience and flexibility of CNF composite foams were poor. All the samples were unable to recover their original shape after compression.

Table 2 Physical property data of CNF composite foams
Full size table

Thermogravimetric analysis (TGA) and differential TGA (DTG) were used to evaluate the thermal stability of pure CNF and CNF composite foams under nitrogen atmosphere. TGA and DTG curves of CNF foams and PA-GZ-CNF composite foams are displayed in Fig. 3. Although the amount of PA-GZ increased from 0 to 25%, the composite foams exhibited a similar trend in TGA curves, the weight of char residue was greatly improved compared with that of pure CNF foam (17.8%). For instance, when the percentage of PA and GZ increased to 25%, the weight percentage of char residue increased to 34.2%, which proved that PA-GZ-CNF composite foams exhibited better carbonization capacity.

Fig. 3
figure 3

a TGA and b DTG curves of CNF foams and CNF composite foams with different PA-GZ contents under nitrogen atmosphere

Full size image

In nitrogen atmosphere, the CNF composite foams were not oxidized but underwent two main thermal degradation steps as observed from DTG curves. There were two main pyrolysis peaks at 266 °C and 320 °C, which were attributed to the thermal pyrolysis of cellulose and further decomposition of char residue (Mamleev et al. 2007; Wang Sanchez-Soto 2015). The weight loss below 150 °C was mainly due to the evaporation of moisture absorbed by the CNF foams. The temperature of the first pyrolysis peak was also affected by the incorporation of PA and GZ. With the addition of PA and GZ varied from 0 to 25%, the first pyrolysis peak temperature decreased from 272 to 258 °C in Fig. 3. Furthermore, the first peak pyrolysis rate declined obviously from 0.45 to 0.26%/°C, while the second peak pyrolysis rate also decreased from 0.70 to 0.55%/°C. The removal of the first pyrolysis peak could be attributed to the that the products generated from the early degradation of PA and GZ led to the degradation of the matrix in advance (Zou et al. 2020a, b). The decomposition of phosphoric acid and guanazole produced some non-flammable gases, which diluted the concentration of combustible gases (Chi et al. 2020).

The LOI and UL-94 vertical burning tests were carried out to analyze the combustion properties of CNF composite foams. As shown in Fig. 4, for pure CNF foam, it was easy to ignite and burned quickly with fast flame propagation. After removing the flame source, it still smoldered for a certain time and completely burned with fragile char residues left. Compared to pure CNF foams, PA-GZ-CNF-3 showed excellent self-extinguishing ability after flame removal. It burned to a shorter length and most of the sample was still intact after exposing to fire for 10 s twice (Fig. 4). All the PA-GZ-CNF composite foams could reach the UL-94 V-0 rating. CNF foam was easy to ignite and its oxygen index was measured to be 24%. After incorporating PA and GZ, the LOI value of PA-GZ-CNF-1 and PA-GZ-CNF-2 was 29% and 31%, respectively. The LOI value of PA-GZ-CNF-3 was increased to 38%. The exceptional flame retardancy of PA-GZ-CNF foams was attributed to the nitrogen-containing and phosphoric compounds produced by PA and GZ molecules during combustion, which promoted the formation of a char layer and covered the matrix surface to inhibit the CNF foams from combustion.

Fig. 4
figure 4

Digital photographs of UL-94 vertical burning tests of a CNF and b PA-GZ-CNF-3, and c the LOI value and UL-94 rating of different samples

Full size image

The combustion behavior of CNF foam and PA-GZ-CNF composite foams was further investigated by a cone calorimetry test (CCT). Many important fire parameters could be obtained from CTT results to simulate fire scenarios, such as total heat release (THR), peak heat release rate (PHRR), total smoke production (TSP), smoke production rate (SPR), and so on. The THR and heat release rate curves of PA-GZ-CNF composite foams are displayed in Fig. 5, while the PHRR, THR, and fire spread index (FGI) values were summarized in Table 3. In Fig. 5a, the pure CNF foam was ignited quickly and showed PHRR and THR values of 57.8 kW/m2 and 2.1 MJ/m2, respectively. The PHRR value decreased gradually with the increase of PA and GZ contents in CNF composite foams. The most significant reduction was observed in PA-GZ-CNF-3 composite foam with a PHRR value (29.27 kW/m2), which corresponded to a 49.4% reduction compared with that of pure CNF foam. As shown in Fig. 5b, there was also a significant decrease in the THR value with the introduction of PA and GZ in CNF foams. For example, the addition of 12.5 wt% PA and 12.5 wt% GZ in the CNF foam brought about a 42.9% reduction in THR, indicating that the presence of PA and GZ effectively enhanced the flame retardant property of CNF composite foams. The fire spread index (FGI) is defined as the ratio of PHRR and time to PHRR (TPHRR), which is an important comprehensive index to evaluate the fire safety of polymer materials. The lower the FGI value, the higher the fire safety. The FGI value of pure CNF was 3.85 kW/m2·s, while PA-GZ-CNF-3 foam showed the lowest FGI value (2.09 kW/m2·s), indicating the highest fire safety. These results showed that PA and GZ greatly reduced the THR and PHRR during the combustion process, and the decomposition of PA and GZ promoted the thermal degradation process of the CNF matrix to form a protective char layer.

Fig. 5
figure 5

a HRR and b THR versus time curves of CNF foam and PA-GZ-CNF composite foams at 35 kW/m2

Full size image
Table 3 Cone calorimeter data of CNF and PA-GZ-CNF composite foams
Full size table

The evolution of gaseous products during the thermal decomposition process of CNF foams was detected by the TG-FTIR technique. The total absorption spectra of the gaseous products and FTIR spectra at the maximum thermal decomposition rate of CNF and PA-GZ-CNF-2 are shown in Fig. 6. It was obvious that the total pyrolysis product for PA-GZ-CNF-2 composite foam was less than that of pure CNF foam (Fig. 6a) and the addition of PA and GZ into CNF restrained the generation of the gaseous products during the combustion process to prevent further combustion. The main gaseous products for CNF composite foams were water (3600–4000 cm−1), CO2 (2307–2380 cm−1) (Wen et al. 2018), hydrocarbons (2800–3100 cm−1) (Wang et al. 2013), CO (2100–2200 cm−1) and carbonyl compound (1600–1800 cm−1). The pyrolysis products for PA-GZ-CNF-2 foam were similar to those of pure CNF foam, indicating that the introduction of PA and GZ did not affect the evolved gaseous products. The change of the decomposition gases versus temperature for CNF composite foams is displayed in Fig. 7 to further describe the change of some typical volatile products. The absorbance intensity of CO, CO2, hydrocarbons, and carbonyl compounds of PA-GZ-CNF-2 foam was lower than that of pure CNF foam. Since most of these volatile products were flammable substances, the decline of flammable organic volatiles decreased the fuel supply and led to a lower HRR value. Meanwhile, PA -GZ-CNF composite foams possessed excellent char formation ability and promoted the formation of a char layer to protect the CNF matrix, which prevented the pyrolysis products from spreading into the gas phase.

Fig. 6
figure 6

a Total absorption spectra and b FTIR spectra of the gaseous products for pure CNF foam and PA-GZ-CNF-2 foam at the maximum decomposition rate

Full size image
Fig. 7
figure 7

Intensity of the gaseous products for pure CNF foam and PA-GZ-CNF-2 foam: a CO2, b CO, c carbonyl compound, and d hydrocarbon

Full size image

The digital photographs of the char residues of CNF composite foams are shown in Fig. 8. It can be observed that there was only thin and fragile residue left for pure CNF foam after the cone calorimeter test, whereas the PA-GZ-CNF composite foams showed a thick and intumescent char residue. Moreover, all the char residues of PA-GZ-CNF composite foams had a white protective barrier covering the surface to protect the internal char, which indicated that the incorporation of PA and GZ could promote the formation of a protective shield and inhibit the combustion process.

Fig. 8
figure 8

Digital photographs of the residues after cone calorimeter test: a CNF, b PA-GZ-CNF-1, c PA-GZ-CNF-2 and d PA-GZ-CNF-3

Full size image

The structure and morphology of char residue are very important for the flame retardant properties of materials. Raman spectroscopy was utilized to explore the chemical structure of char residues of the CNF composite foams. Figure 9 shows the Raman spectra of the char residues of CNF composite foams and the Raman spectra curves of each sample exhibited two main bands at 1360 and 1600 cm−1, which were ascribed to the D-band and G-band. The D-band is related to the amorphous carbon atoms which reflects the sp3 defects in carbon materials, while the G-band is relevant to the graphited carbon atoms that reflect the vibration of graphitic carbon atoms (Shi et al. 2020; Xu et al. 2021). So, the intensity ratio of the D-band and G-band (ID/IG) is used to estimate the graphitization degree of materials (Huang et al. 2021; Ji et al. 2020). The higher ID /IG value indicates the higher defect degree and the lower graphitization degree. The ID/IG value of the CNF composite foams was lower than that of pure CNF foam (2.97), indicating a higher graphitization degree. With the increased content of PA-GZ in the samples, the value of ID/IG gradually decreased from 2.97 to 1.88. The char residue of the PA-GZ-CNF foams included a higher graphitization structure, resulting in better thermal resistance of the residual chars under high temperatures.

Fig. 9
figure 9

Raman spectra of the residual chars of CNF and PA-GZ-CNF composite foams

Full size image

In order to further characterize the element distribution of the char residue, XPS analysis was utilized to investigate the char residue of CNF composite foams obtained after the cone calorimeter test. The XPS survey spectra of the char residues for pure CNF foam showed two intense signals of C and O, and the additional signals of N and P were also observed in PA-GZ-CNF composite foams in Fig. 10. In the case of the C1s spectrum of pure CNF foam, the peaks at 288.4, 286.4, and 284.8 eV were assigned to C=O, C–O, and C–C (Fig. 11a) (Guo et al. 2019; Shi et al. 2018a, b). Different from that, the C1s spectrum of PA-GZ-CNF foams showed two new peaks at 286.0 and 285.6 eV, which belonged to C–N and C–P bonds due to the introduction of PA and GZ into cellulose (Fang et al. 2020). Similar to the C1s spectrum, the O1s spectra presented two new peaks at 530.9 and 533.1 eV which were allocated to O–P and O=P compared with pure CNF foam (Fig. 11b) (Shi et al. 2018a, b). As shown in Fig. 11c, two kinds of nitrogen configurations appeared at 400.1 eV and 398.6 eV, corresponding to N–C and N=C structures in guanazole molecules (Wang et al. 2019). Moreover, in the P2p spectrum, the peaks at 132.7, 133.4, and 134.05 eV were corresponding to the P–O, P=O, and P–C groups in phosphorus‐rich structures of char residues (Hou et al. 2018; Zou et al. 2020a, b). The thermal-oxidative resistance of char residues depended on the content of unoxidized carbon atoms like aliphatic and aromatic carbons in high‐resolution C1s spectra (Bourbigot. et al. 1997; Liao et al. 2012). So the ratio between Co (Co: oxidized carbons) and Cuo (Cuo: unoxidized carbons) was utilized to evaluate the thermal-oxidative resistance of materials (Wang et al. 2013). After calculation, the ratio value of Co/Cuo of PA-GZ-CNF-2 foam was 0.58, which is much lower than that of pure CNF foam (1.45), which meant higher content of the unoxidized carbons in the char residues of composite foams compared with pure CNF foam. Thus, the char residue owned higher thermal stability and protected the internal matrix from oxidation during combustion.

Fig. 10
figure 10

XPS survey spectra of CNF and PA-GZ-CNF-2

Full size image
Fig. 11
figure 11

a C1s and b O1s XPS spectra of CNF and PA-GZ-CNF-2; c N1s and d P2p XPS spectra of PA-GZ-CNF-2

Full size image

Cellulose is a biopolymer that consists of glucose residues, and has a high density of hydroxyl groups in its molecular structure, and forms a strong affinity with water. Thus, hydrophobic treatment is required for cellulose to satisfy various practical applications (Chen et al. 2018; Ferreira et al. 2020). There are several methods to modify cellulose for hydrophobic treatment, including hexafluoropropene plasma treatment (Li and Dai 2007), tetramethylsilane (Kwong et al. 2014), metal oxide nanoparticles (Dolez et al. 2017), and silane coupling agents (Siy et al. 2020). In this work, polydimethylsiloxane (PDMS) was chosen to modify the surface of cellulose composite foams. Due to its intrinsic hydrophilic property, the water droplets were absorbed on the surface of pure CNF foam with a contact angle of 0°. PA-GZ-CNF-3 was chosen for hydrophobic treatment and named PDMS-PA-GZ-CNF-3. Compared with that of pure CNF foam, the water droplet could retain a fairly complete shape with a contact angle of 104.0° on the surface of PDMS-PA-GZ-CNF-3 (Fig. 12), indicating that the CNF composite foams successfully achieved a certain hydrophobicity. The stress at 80% strain of PDMS-PA-GZ-CNF-3 composite foam (81.01 kPa) was much higher than that of pure CNF foam. Moreover, CNF composite foams can also maintain remarkable flame retardancy and pass the UL-94 V-0 rating (Fig. 13a). The XPS survey spectra of the char residue for PDMS PA-GZ-CNF-3 showed a new signal of silicon (Fig. 13b). In the case of the Si2p spectrum, the peaks at 103.8 and 102.4 eV were assigned to Si–Si/Si–C and Si–O formed during the combustion process (Chen et al. 2020), which indicated that the combination of PDMS not only enhanced the hydrophobicity but also promoted the formation of a more compact char layer to prevent combustion.

Fig. 12
figure 12

Digital photographs of contact angle tests a CNF and b PDMS-PA-GZ-CNF-3 foams

Full size image
Fig. 13
figure 13

a Digital photographs of PDMS-PA-GZ-CNF-3 composite foam in the UL-94 vertical burning test; b XPS survey spectra and c Si2p XPS spectra of PDMS-PA-GZ-CNF-3 composite foam

Full size image

Conclusion

In summary, we combined renewable cellulose nanofibers (CNFs) with phytic acid (PA) and guanazole (GZ) to prepare flame retardant CNF composite foams. The CNF composite foams showed a highly hierarchical porous structure, remarkably increased compression strength, and excellent flame retardancy. The LOI value of PA-GZ-CNF-3 foam was 38%, much higher than that of pure CNF foam (24%). PA-GZ-CNF composite foams exhibited self-extinguishing ability in the UL-94 vertical burning test and exhibited better carbonization capacity compared with pure CNF foam in TGA results. Based on the cone calorimeter tests, it possessed a very low peak heat release rate (29.27 kW/m2) and total heat release (1.21 MJ/m2), indicating the significantly enhanced flame retardancy. The presence of PA and GZ decomposed earlier and promoted the degradation process of the CNF matrix to form a protective char layer during combustion. Moreover, PDMS-PA-GA-CNF-3 composite foam also retained great flame retardancy and passed the UL-94 V-0 vertical combustion test after hydrophobic treatment, which showed hydrophobicity with a contact angle of 104.0°.