Effect of Sepiolite-Loaded Fe2O3 on Flame Retardancy of Waterborne Polyurethane

Key Laboratory of Science & Technology of Eco-Textile, Ministry of Education, Donghua University, Shanghai 201620, China Tsinghua Univ, Dept Chem, Minist Educ, Key Lab Bioorgan Phosphorus Chem & Chem Biol, Beijing 100084, China Zhejiang University, Institute for Composites Science Innovation, Hangzhou, Zhejiang Province, 310058, China National Dyeing and Finishing Engineering Technology Research Center, Donghua University, No. 2999, North Renmin Road, Songjiang District, Shanghai 201620, China National Manufacturing Innovation Center of Advanced Dyeing and Finishing Technology, Taian, Shandong Province 271000, China


Introduction
As an eco-environmentally polymer material that could replace traditional solvent-based polyurethane, waterborne polyurethane (WPU) has been used more and more widely in producing apparel, clothing, home furnishings, and industrial products in recent years [1,2]. However, the flammability and poor heat resistance of WPU are a huge obstacle to its application in firefighter's protective clothing and military and airline industry [3]. Previous research on flame retardant polyurethane has mainly focused on solvent-based polyurethane (PU) [4,5]. For example, Chen et al. studied on thermal properties of PU nanocomposites based on organic sepiolite [6], but the thermal stability of PU composites is not as good as expect. Zheng et al. researched the characterization of sepiolite/PU nanocomposites [7]; the TG results of sepiolite/PU nanocomposites identified slightly improved. Due to the different dispersion systems of WPU and PU, their molecular structures and properties are also different, leading to the requirements for flame-retardant property also being different. The current situation is that there are a few of studies on the flame-retardant property mainly for WPU. Han et al., Kim et al., and Yeh et al. studied the flameretardant property of WPU/clay composites; the flameresistant property of these materials could only be improved a little [8][9][10].
As society is progressing, human beings become more aware of environmental protection. Even though many conventional flame retardants exhibit excellent flame resistance, halogen-containing flame retardants are banned due to serious environmental problems. Phosphorus-and nitrogen-containing flame retardants as the second category of optional flame retardants still have some environmental problems, such as eutrophication of water caused by the discharge of phosphorus-containing wastewater into the natural world [11][12][13][14][15][16][17][18][19][20][21]. Therefore, there is an urgent need for an environmentally friendly, efficient, and convenient flame retardant.
The researchers proposed that sepiolite-(Sep-) loaded metal oxides could be used for preparing flame-retardant polymer. Sep as a natural silicate material, which is not combustible itself, is nonpolluting to the environment and nonhazardous to human beings. However, when the material is used as flame retardant alone, the flame retardancy is not excellent [22][23][24]. For these reasons, an inorganic composite flame retardant is prepared by loading the transition metal oxide which has Lewis's acid property and could act as a kind of catalytic carbon formation agent on Sep. Therefore, this inorganic composite flame retardant could enhance the flame-retardant property of polymer by improving the carbon formation of the substrates. For example, Guillon prepared a series of catalyst which comprised at least one IZM-2 zeolite, one matrix, and one metal selected from metals of groups VIII, VIB, and VIIB. Those catalysts promote rapid thermal decomposition of polymers to generate carbonization [25]. Aranda et al. prepared layered double hydroxide/Sep heterostructured materials and investigated the catalytic properties of the above materials [26]. Yuming Chen and Limin Zhou studied energetics and mechanisms of graphene growth catalyzed by nickel nanoclusters by using ab initio density functional theory calculations [27].
These previous works made believed that loading transition metal oxides into Sep could enhance the carbonforming effect of Sep. However, what will happen if transition metal oxide is combined with Sep to form a flameresistant system for WPU? In this study, inorganic composites (Sep@Fe 2 O 3 ) were prepared by combining precipitation and calcination methods to load Fe 2 O 3 into Sep crystals; organic/inorganic composites were prepared by combining these inorganic composites with WPU by layer-by-layer (LBL) method to study the thermal stability and combustion properties of these composites and explore the mechanism of the effect of Sep@Fe 2 O 3 on the combustion property of WPU. atmosphere for 45 min. In this system, the amount of iron ions is double the cation exchange capacity (CEC) of Sep (93.7 mmol/100 g). Then, 100 mL NaOH solution (0.11 mol/L) was added to the solution with high-speed stirring for 2 h. The slurry aged in the darkroom for 20 min and then was separated by centrifugation (TG20G, Spring Instrument. Co., Ltd., China), washed by deionized water, and dried at 105°C for 4 h. The prepared dry chips were calcinated in muffle furnace (GST-3-1200-I, Shanghai Guangshu Electromechanical Co., Ltd.) at 400°C for 4 h to prepare Sep@Fe 2 O 3 composite material.

Experimental
2.3. Preparation of WPU/Sep@Fe 2 O 3 Composites. All films were assembled on a given substrate using the procedure shown in Figure 1. Each substrate was dipped into the deposition mixtures, alternating between the WPU emulsion of 3.0 wt.% and inorganic materials (Sep, Fe 2 O 3 , and Sep@-Fe 2 O 3 ) of 0.2 wt.%, with each cycle corresponding to one bilayer (BL). Each dip was followed by handling with a stream of filtered air for 60 s, the above process was repeated for 10 times, and then WPU composites were prepared. Samples of WPU composites are shown in Table 1.

Characterization.
The microstructure characterization of Sep samples and WPU composites was carried out by X-ray diffractometer (XRD, 6100), transmission electron microscope (TEM, SU-70t), and scanning electron microscopy (SEM, SU-3500). The XRD spectrum of synthesized catalysts was obtained by radiation Cu Ka (λ = 1:54 Å). Samples of TEM were embedded in 812 resin (TAAB, UK), and ultrathin sections (about 800 nm thick) were cut on a Leica ultramicrotome (Leica Microsystems, Milton Keynes, UK sections) viewed in a SU-70t TEM. The election high tension (EHT) of SEM was set at 10 kV, and imaging was carried out using SE2 detector. Due to the nonconductive nature of the WPU composites, the samples were sputter-coated with a thin layer of gold prior to the imaging using a Polaron SC 7640 for 60 s. ATR-FTIR spectrophotometry was used to determine the surface structure of the WPU composites and to qualitatively determine the presence of the organic modifiers in the WPU composites. A Tensor 27 model infrared spectrophotometry with attenuated total reflectance module with monolithic diamond crystal for full spectral range from 4000 to 400 cm -1 was used.
The thermal stability of samples was measured in a thermal gravimetric analyzer (TG, SDT Q600). Each sample was approximately 5 mg and tested from 30°C to 900°C at a heating rate of 20 mL min -1 . The vertical flame test (VFT) was performed according to BG/T 5455-1997, by using a vertical burning tester (BCR-A, Textile Science Institute, Shandong, China). The samples (300 mm × 80 mm) were prepared using the procedure shown in Figure 1. The abovementioned substrate for preparing the WPU composite film was changed into cotton textile (115 g/m 2 , Zhuocheng Special Textile Co. Ltd., Xinxiang, China), and we could get cotton-WPU composite materials. Then, the samples held 19 mm over the Bunsen burner were first exposed to the flame for a period of 12 s and then removed rapidly, the test was repeated three times for each sample, and after which, 2 Advances in Polymer Technology the after-flame and after-glow times were read from the VFT machine.
The combustion performance of samples was determined by a Microscale Combustion Calorimeter (MCC, MCC-2, Govmark, USA) instrument equipped with a 40 μL alumina pan. The samples (~5 mg) were first heated from room temperature to 750°C in an 80 cm 3 /min nitrogen stream flowing at a linear heating rate of 1°C. Then, the gaseous pyrolysate mixture was mixed with a 20 cm 3 /min oxygen stream flow and combustion in a furnace at 900°C for 10 s.
Tensile tests of WPU composites were carried out according to standard GB/T 1040.3-2006 on UTM 6203 machine (Shenzhen Sansi Vertical and Horizontal Technology Co., Ltd., China), and the test was repeated 3 times for each sample.  Figure 2. In the XRD patterns of the Sep and Sep@Fe 2 O 3 composites, the characteristic peaks of Sep were weak in the Sep@Fe 2 O 3 composites, indicating that the addition of Fe 2 O 3 in the Sep@Fe 2 O 3 composites affects the reading of the characteristic peak data of Sep [28]. The peak positions of Sep@Fe 2 O 3 composites show good agreement to the standard data of Fe 2 O 3 structure (PDF: 85-0987, 24.20°represents the characteristic peak of the 110 crystal plane, 35.73°represents that of -110 crystal plane). We observe reflexes associated with the (110) and (-110) planes; the intensity of the (110 crystal plane) reflexes is the highest. This suggests the successful preparation of Sep@Fe 2 O 3 composites. In addition, no characteristic peak of other materials is found in XRD patterns, suggesting that the material is the composites of Sep with Fe 2 O 3 .

Result and Discussion
The results of crystal characterization of WPU film and WPU/Sep@Fe 2 O 3 film are shown in Figure 2. XRD diffraction pattern of WPU film shows a broader diffraction peak, indicating that the WPU film is semicrystalline in nature. The peak at about 22.56°, which is the characteristic peak of WPU, corresponds to the (110) plane of WPU representing scattering from WPU chains at interplanar spacing   3 Advances in Polymer Technology [29][30][31]. In addition, the XRD of pure WPU shows the low crystallinity of the sample which is due to repetition of benzenoid rings [32]. The

Advances in Polymer Technology
Sep framework, and then, the Sep could convert to porous silica [33]. In this research, part of Fe 3+ is incorporated into the magnesium vacancies of Sep to form variety of stable metalliccontaining Sep [34,35]. Once all the magnesium vacancies have been occupied and in case the aqueous solution contains an excess of Fe 3+ , which would precipitate on the Sep surface as crystallized Fe 2 O 3 or Fe (OH) 3 , this is the case of metal ion interacting via electrostatic attraction. The Sep microfiber is an anionic clay mineral that offers binding sites to immobilize the positively charged particles through electrostatic interactions. In addition, from the TEM image (Figure 4(b)), we could analyze that the diameter of Fe 2 O 3 is about 25-70 nm by ImageJ (shown in Figure 5 (Figure 4(e)) shows that the particle size of Fe 2 O 3 crystal is not uniform (shown in nanoscale and micron scale, Figure 5(b)). In addition, there is no obvious phase interface between Fe 2 O 3 crystal and WPU matrix. This could prove that the compatibility of organic and inorganic phases is very excellent. TEM of WPU/Sep@Fe 2 O 3 composite membrane is shown in Figure 4(f), in which the Fe 2 O 3 crystal loaded abundantly on the surface of Sep fibers, which dispersed in the WPU matrix. No phase interface could be found between the interface of Sep and WPU matrix; it means that the two phases (organic and inorganic phases) are compatible. This is manifested in the greater improvement in mechanical and thermal properties which would be discussed in the subsequent sections.

Advances in Polymer Technology
flame retardant. Therefore, Sep, Fe 2 O 3 , and Sep@Fe 2 O 3 were used to prepare composites with WPU by LBL method. The TG and DTG thermograms of WPU and WPU composites were analyzed under air atmosphere shown in Figures 6(a) and 6(b). The first stage at 200-300°C involves the elimination reactions of H 2 O, CO 2 , and NH 3 and residual acetate groups because of partially hydrolyzing of WPU; the degradation step at 300-600°C is more complex and includes the further degradation of WPU residues to yield the carbon and hydrocarbons shown in Figure 6(a). The hydrocarbons in the TGA results have been discussed by our previous studies. The weight loss of WPU/Sep, WPU/Fe 2 O 3 , and WPU/Sep@Fe 2 O 3 composites took place at 300-400°C and 400-600°C, for the loss of structure water of Sep and the thermal decomposition of WPU molecular. As is shown in Figure 6(a), at 600°C, the carbon residual percentages of WPU/Sep, WPU/Fe 2 O 3 , and WPU/Sep@Fe 2 O 3 composites are 7.3%, 12.2%, and 13.4%, respectively, showing obvious difference from the control sample (1.4% for WPU). This increased percentage of char residue implies that Sep, Fe 2 O 3 , and Sep@Fe 2 O 3 are favorable to improve the hightemperature resistance of WPU. Sep could act as the mass transport barrier to the volatiles produced during decomposition. Physi-and/or chemisorption occurring in the filler (Sep and Fe 2 O 3 )-matrix interphase may lead to the formation of physical network thereby enhancing the nanocomposite stability [7,36].
The DTG (%/min) curves for thermal oxidation of the representative WPU and WPU composite samples are provided in Figure 6(b). Almost all the samples displayed two degradation peaks. It is noteworthy that the temperature at the greatest degradation peaks (T max ) decreased from    film) were burned just for 10 s and retained the original fiber structure of the textiles. Shown in Figure 7 are SEM images of cotton residues. From Figures 7(a)  In the presence of the flame-retardant system, the HRR peaked at 355.2°C for the WPU film (shown in Table 3). It increased by 30°C and decreased by 52°C to become two peaks at 385°C (269.9 W/g) and 303°C (241.9 W/g) for the WPU/Sep composite film. As is shown in Table 3, the total heat release (THR) of WPU decreased from 27 to 26 kJ/g whereas the percent char yield of WPU increased from 3.1 to 6.9% after being complexed with Sep. Moreover, the one peak in the HRR curve of the WPU split into two peaks after being complexed with Sep, and the HRR curve of the WPU/ Sep composite film became broader with a less symmetric shape than that of WPU. When 6.7 wt.% Sep was replaced by 6.7 wt.% Fe 2 O 3 or 6.7 wt.% Sep@Fe 2 O 3 , peak of heat release rate (pHRR) and THR decreased, and the temperature of heat release rate (T PHRR ) has little change. These results demonstrate the formation of a char layer over the surface of the material during the combustion process; the char layer could inhibit gaseous-fuel transport from the WPU to the flame front so as to reduce the HRR of the burning surface according to the combustion mechanism of clay-based composites during the MCC experiments. In addition, the char yield increased from 3.1% (for WPU) to 6.9%, 7.6%, and 14.3% for WPU/Sep, WPU/Fe 2 O 3 , and WPU/Sep@ Fe 2 O 3 , respectively, indicating that the addition of inorganic materials could help the WPU to retain more degradation products in the condensed phase during combustion. The data suggested that the flame-retardant property of Sep@-Fe 2 O 3 is better than that of Sep or Fe 2 O 3 alone. In addition, the results of MCC are coincident with the results of TG test (Figure 6(a)).

Tensile
Properties. The mechanical properties of WPU and WPU composites are shown in Table 4. The increasing of breaking strength supports the strengthening of mechanical strength, and the decreasing of break elongation indicates the weakening of elasticity of WPU composites. The data presented in the table are the average of three readings. The breaking strength of S-2 decreased from 3.2 N to 2.7 N, as the result of stress concentration due to agglomeration of Sep microfibers. In the case of S-3, breaking strength increased from 3.2 N to 7.0 N, because the addition of Fe 2 O 3 into the WPU matrix improved the interfacial adhesion among the WPU matrix through oxygen atoms (in Fe 2 O 3 structure) and amino and hydroxyl groups (in WPU molecular structure), respectively [38,39]. Considering Figure 4(e), the interface between Fe 2 O 3 and WPU could not be observed, which means that the compatibility between those two materials is excellent. Therefore, the result of tensile properties of WPU composites is consistent with that of TEM. Sample S-4 was prepared by WPU and Sep@Fe 2 O 3 , which showed a significant improvement in mechanical property and small reduction in elasticity property. For oxygen atom and hydroxyl in Sep matrix and -NH-, C=O in WPU could form hydrogen bond formation to improve the mechanical strength of WPU composites. Meanwhile, interaction between WPU and inorganic materials could restrict the movement of polymer chains and reduce the elasticity property of WPU composites. In summary, the heat resistance and mechanical properties of WPU/Sep@Fe 2 O 3 composites are confirmed in this work, proving that it has the potential to be used as a high temperature-resistant material.

Conclusion
In this work, Fe 2 O 3 as a kind of flame retardant is loaded in Sep that was used for WPU. Flame-retardant properties of Sep, Fe 2 O 3 , and Sep@Fe 2 O 3 were evaluated by VFT, TGA, and MCC measurements. The compounding of Sep@Fe 2 O 3 with WPU could change the thermal stability of the material. Using VFT, it was found that the flame-retardant property achieved optimization when 6.7 wt.% Sep@Fe 2 O 3 substituted 6.7 wt.% Sep or Fe 2 O 3 . In this case, the residual carbon percentages of WPU, WPU/Sep, WPU/Fe 2 O 3 , and WPU/Sep@-Fe 2 O 3 composites obtained from TG/DTG curves are 1.4%, 7.3%, 12.2%, and 13.4%, respectively. These results revealed that the thermal stability of WPU/Sep@Fe 2 O 3 increased compared with that of WPU/Sep or WPU/Fe 2 O 3 . This phenomenon can also be observed in the MCC tests in terms of

Data Availability
The data is included in the manuscript already.

Conflicts of Interest
The authors declared that there are no conflicts of interest in this work. We declare that we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.