Bio-Based Rigid Polyurethane Foams Modified with C-MOF/MWCNTs and TBPBP as Building Insulation Materials: Synergistic Effect and Corresponding Mechanism for Enhancing Fire and Smoke Safety

Rigid polyurethane foams (RPUFs) as building insulation materials quickly burn and release a lot of heat, smoke, and carbon monoxide, and cause human safety risk and severe environmental pollution. To mitigate these disadvantages, MOF/MWCNTs were fabricated via mixing Cu ions’ partly substituted framework of ZIF-67 and MWCNTs, and further calcinated MOF/MWCNTs (C-MOF/MWCTs) was newly generated by calcinating MOF/MWCNTs in air. Then, MOF/MWCNTs and C-MOF/MWCNTs were respectively employed together with a phosphorus–nitrogen-containing reactive flame retardant (TBPBP) to prepare renewable bio-based rigid polyurethane foam, including RPUF-T/MOF/MWCNTs 2 and RPUF-T/C-MOF/MWCNTs 2. The characterization results showed that RPUF-T/C-MOF/MWCNTs 2 had better performance than RPUF-T/MOF/MWCNTs 2 and neat RPUF. Compared to neat RPUF, the compressive strength, limiting oxygen index value, and the mass char residue in cone calorimetry test of RPUF-T/C-MOF/MWCNTs 2, respectively, were increased by 105.93%, 46.35%, and 347.32%; meanwhile, the total heat release rate, total smoke production, total carbon monoxide product, and total carbon dioxide product were reduced by 47.97%, 50.46%, 41.38%, 43.37%, respectively. This study provides a referable method for preparing RPUFs with good physical properties, fire, and smoke safety, which is favorable for human safety and environmental protection as new building insulation materials.


Introduction
Bio-based polyurethanes (PUs) mostly derived from renewable bio-based polyols or isocyanates are beneficial to their development in view of environmental protection and sustainability issues [1]. Among bio-based PUs, bio-based rigid polyurethane foam (RPUFs) accounts for a large proportion because they possess high specific strength, low heat conductivity, light weight, and low cost, resulting in their widespread use in various fields as thermal insulation materials [2,3]. Nonetheless, bio-based RPUFs burn easily and release a lot of heat, smoke, and carbon monoxide during their combustion. In order to protect human property and life safety and our environment, it is critical to endow bio-based RPUFs with flame-retardant and smoke-suppressing properties.

Preparation of MOF/MWCNTs and C-MOF/MWCNTs
In total, 1 g MWCNTs, 3 g PVP K30, and 500 mL absolute methanol were added into a 1000 mL glass beaker, then the glass beaker was placed in an ultrasonic cleaner with continuous ultrasonic vibration for 60 min, which made MWCNTs and PVP K30 evenly dispersed in absolute methanol. Then, 4.832 g Cu(NO3)2·3H2O and 11.640 g Co(NO3)2·6H2O were doped into the above beaker with continuous ultrasonic vibration for 60 min. Next, 49.284 g 2-MIM was dispersed into 200 mL absolute methanol in a 500 mL glass beaker. Then, the solution in the 500 mL beaker was poured into the 1000 mL beaker above. The 1000 mL beaker was placed on magnetic stirrer and stirred at a rate of 400 rpm for 20 min. The 1000 mL beaker loaded with the above reagents was placed in a stable place and left to stand for 10 days. Blue-purple metal organic frameworks mixing with multi-walled carbon nanotubes (MOF/MWCTs) were obtained by filtration and drying. Absolute methanol was again gained via rotating evaporation and reused in the next preparation of MOF/MWCTs. Prepared MOF/MWCTs were transferred into a muffle furnace so as to calcine at 500 °C for 180 min in air. Black calcined MOF/MWCTs (denoted as C-MOF/MWCTs) were obtained. The synthetic route of MOF/MWCTs and C-MOF/MWCTs were presented in Scheme 1.

Preparation of Bio-Based Rigid Polyurethane Foams (RPUFs)
Apart from pMDI, the remaining reagents in Table 1 were added to a plastic beaker and stirred at the rate of 400 rpm for 10 min. Subsequently, the corresponding mass of pMDI was swiftly poured into the plastic beaker when the stirring rate was 1100 rpm. After about 10 s, the components in the beaker were rapidly poured onto a steel plate at 60 °C for free foaming. After ca. 3 min, the steel plate was transferred to a vacuum drying oven and dried at 60 °C for 10 h. Bio-based RPUFs were fabricated and then cut down to the appropriate size for various tests. The synthesis diagram of RPUF-T/C-MOF/MWCNTs 2 was presented in Scheme 2. Other RPUFs were prepared via utilizing similar preparation processes to that of RPUF-T/C-MOF/MWCNTs 2.

Preparation of Bio-Based Rigid Polyurethane Foams (RPUFs)
Apart from pMDI, the remaining reagents in Table 1 were added to a plastic beaker and stirred at the rate of 400 rpm for 10 min. Subsequently, the corresponding mass of pMDI was swiftly poured into the plastic beaker when the stirring rate was 1100 rpm. After about 10 s, the components in the beaker were rapidly poured onto a steel plate at 60 • C for free foaming. After ca. 3 min, the steel plate was transferred to a vacuum drying oven and dried at 60 • C for 10 h. Bio-based RPUFs were fabricated and then cut down to the appropriate size for various tests. The synthesis diagram of RPUF-T/C-MOF/MWCNTs 2 was presented in Scheme 2. Other RPUFs were prepared via utilizing similar preparation processes to that of RPUF-T/C-MOF/MWCNTs 2. Note: 1. Neat RPUF and RPUF-T, respectively, refer to neat RPUF and RPUF-T45 in our previous work [11]. 2. Neat RPUF and RPUF-T are introduced to systematically study the effects of MOF/MWCNTs and C-MOF/MWCNTs on the physical properties, flame retardancy, and smoke suppression of bio-based RPUFs. 3. The unit of all reagents is gram. 154.91 Note: 1. Neat RPUF and RPUF-T, respectively, refer to neat RPUF and RPUF-T45 in our previous work [11]. 2. Neat RPUF and RPUF-T are introduced to systematically study the effects of MOF/MWCNTs and C-MOF/MWCNTs on the physical properties, flame retardancy, and smoke suppression of bio-based RPUFs. 3. The unit of all reagents is gram.

Characterization
Field emission scanning electron microscope (FSME) images and energy dispersive X-ray spectroscopy (EDXS) images of MOF/MWCTs and C-MOF/MWCTs, respectively, were gained by a field emission scanning electron microscope (ZEISS sigma500, Carl Zeiss AG, Oberkohen, Germany) with energy dispersive X-ray spectrometer (Bruker XFlash 6130, Bruker Corporation, Karlsruher, Germany) at the voltage of 10 kV. FSME images of the microstructures of the section and char residues of RPUFs were attained via field emission scanning electron microscopes (Sirion 200, FEI Company, Hillsboro, OR, USA) and (EM-30 Plus, COXEM Co., Ltd., Daejeon City, Korea).
Transmission electron microscopy (TEM) images and energy dispersive X-ray spectroscopy (EDXS) images of MOF/MWCTs, C-MOF/MWCTs, and char residues of RPUFs, respectively, were obtained via a transmission electron microscope (JEOL-JEM 2100 F, JEOL, Showa, Tokyo, Japan) with energy dispersive X-ray spectrometer at the voltage of 80 kV. Before the TEM test, MOF/MWCTs and C-MOF/MWCTs were respectively added into alcohol, and then the mixture was ultrasonic in an ultrasonic cleaning machine.

Characterization
Field emission scanning electron microscope (FSME) images and energy dispersive X-ray spectroscopy (EDXS) images of MOF/MWCTs and C-MOF/MWCTs, respectively, were gained by a field emission scanning electron microscope (ZEISS sigma500, Carl Zeiss AG, Oberkohen, Germany) with energy dispersive X-ray spectrometer (Bruker XFlash 6130, Bruker Corporation, Karlsruher, Germany) at the voltage of 10 kV. FSME images of the microstructures of the section and char residues of RPUFs were attained via field emission scanning electron microscopes (Sirion 200, FEI Company, Hillsboro, OR, USA) and (EM-30 Plus, COXEM Co., Ltd., Daejeon City, Korea).
Transmission electron microscopy (TEM) images and energy dispersive X-ray spectroscopy (EDXS) images of MOF/MWCTs, C-MOF/MWCTs, and char residues of RPUFs, respectively, were obtained via a transmission electron microscope (JEOL-JEM 2100 F, JEOL, Showa, Tokyo, Japan) with energy dispersive X-ray spectrometer at the voltage of 80 kV. Before the TEM test, MOF/MWCTs and C-MOF/MWCTs were respectively added into alcohol, and then the mixture was ultrasonic in an ultrasonic cleaning machine.

Characterization of C-MOF/MWCNTs
Physical and chemical properties of C-MOF/MWCNTs were tested before it was incorporated into RPUFS. Fourier-transform infrared (FT-IR) spectrum, X-ray diffraction (XRD) curve, and Raman curve of C-MOF/MWCNTs were exhibited in Figure 1a-c, respectively. Figure 1a shows two very strong characteristic peaks at 568 cm −1 and 661 cm −1 , which were respectively ascribed to the stretching vibration of Co(III)-O and Co(II)-O of Co 3 O 4 [12,13] that was a calcined product of MOF in air. As can be seen from Figure 1b [6,14]. As shown in Figure 1c, the five typical peaks at around 201 cm −1 , 485 cm −1 , 524 cm −1 , 622 cm −1 , and 693 cm −1 were consistent with F 2g (3) , Eg, F 2g (2) , F 2g (1) , and A 1g modes of Co 3 O 4 [15]. In fact, CuO was theoretically created via calcining MOF in air. Interestingly, no obvious characteristic peaks of CuO were observed in Figure 1a-c, which were presumably attributed to low fraction of CuO and amorphous CuO. 77.33°, and 78.40° belonged to the plane diffraction patterns ((111), (220), (311), (222), (400), (422), (511), (440), (533), (622)) of Co3O4 [6,14]. As shown in Figure 1c, the five typical peaks at around 201 cm −1 , 485 cm −1 , 524 cm −1 , 622 cm −1 , and 693 cm −1 were consistent with F2g (3) , Eg, F2g (2) , F2g (1) , and A1g modes of Co3O4 [15]. In fact, CuO was theoretically created via calcining MOF in air. Interestingly, no obvious characteristic peaks of CuO were observed in Figure 1a-c, which were presumably attributed to low fraction of CuO and amorphous CuO.  Figure S2 in Supplementary Materials [16][17][18][19][20][21][22][23][24]), the peak of N1s vanished in Figure 2a, which was assigned to the thermal decomposition of 2-MIM of MOF/MWCNTs in the period of calcination in air. The proportion of specific elements of MOF/MWCNTs and C-MOF/MWCNTs were shown in Table S3. As shown in Table S3, the proportion of C and N atoms decreased, and the proportion of O, Co, and Cu atoms increased. The proportion of C atoms did not decrease significantly, and still accounted for 70% of the atomic proportion, which was caused by the large amount of MWCNTs contained in C-MOF/MWCNTs, because the mass loss of MWCNTs almost did not occur when calcined in air atmosphere of 500 °C. The reason for the increase in the proportion of O atoms was that MOF reacted with oxygen to form CO3O4 and CuO during the  Figure S2 in Supplementary Materials [16][17][18][19][20][21][22][23][24]), the peak of N1s vanished in Figure 2a, which was assigned to the thermal decomposition of 2-MIM of MOF/MWCNTs in the period of calcination in air. The proportion of specific elements of MOF/MWCNTs and C-MOF/MWCNTs were shown in Table S3. As shown in Table S3, the proportion of C and N atoms decreased, and the proportion of O, Co, and Cu atoms increased. The proportion of C atoms did not decrease significantly, and still accounted for 70% of the atomic proportion, which was caused by the large amount of MWCNTs contained in C-MOF/MWCNTs, because the mass loss of MWCNTs almost did not occur when calcined in air atmosphere of 500 • C. The reason for the increase in the proportion of O atoms was that MOF reacted with oxygen to form CO 3 O 4 and CuO during the calcination in air atmosphere. The XPS spectrum of Co2p showed two characteristic peaks and two satellite peaks [25,26]. The peak of Co2p 3/2 was divided into two peaks with the binding energies of 778.5 eV and 780.6 eV for Co 3+ 2p 3/2 and Co 2+ 2p 3/2 species, respectively. The peak of Co2p1/2 was divided into two peaks at around 793.4 eV and 795.5 eV, respectively, which belong to Co 3+ 2p 1/2 and Co 2+ 2p 1/2 species. The results of Figure 2b displayed the coexistence of Co 3+ and Co 2+ in C-MOF/MWCNTs. The Cu2p XPS spectrum (Figure 2c) exhibited four peaks at near 933.8 eV, 942.6 eV, 953.8 eV, and 953.6 eV, which were assigned to Cu2p 3/2 , the satellite peak of Cu2p 3/2 , Cu2p 1/2 , and the satellite peak of Cu2p 1/2 [15]. Evidently, compared with the Cu2p XPS spectrum of MOF/MWCNTs ( Figure S2c), the Cu2p XPS spectrum of C-MOF/MWCNTs fit more precisely because of the increased atomic ratio of Cu via the calcination of MOF/MWCNTs. respectively, which belong to Co 3+ 2p1/2 and Co 2+ 2p1/2 species. The results of Figure 2b displayed the coexistence of Co 3+ and Co 2+ in C-MOF/MWCNTs. The Cu2p XPS spectrum ( Figure 2c) exhibited four peaks at near 933.8 eV, 942.6 eV, 953.8 eV, and 953.6 eV, which were assigned to Cu2p3/2, the satellite peak of Cu2p3/2, Cu2p1/2, and the satellite peak of Cu2p1/2 [15]. Evidently, compared with the Cu2p XPS spectrum of MOF/MWCNTs ( Figure  S2c), the Cu2p XPS spectrum of C-MOF/MWCNTs fit more precisely because of the increased atomic ratio of Cu via the calcination of MOF/MWCNTs. The microstructure of C-MOF/MWCNTs was detected via field emission scanning electron microscope (FSME), and the corresponding images at various magnifications were presented in Figure 3. Compared with FSEM images of MOF/MWCNTs ( Figure S3), C-MOF/MWCNTs maintained the fundamental frame structure of MOF/MWCNTs. In detail, MWCNTs still existed after calcination, and calcined MOF retained the normal dodecahedron structure, whereas the surface of MOF altered from firm to porous because of the decomposition of the organic linker during calcination [26]. FSEM image of C-MOF/MWCNTs with mapping mode and its specific energy dispersive X-ray spectroscopy (EDXS) elemental mapping were shown in Figure S12. The mass ratio and atom ratio of elements of FSEM image of C-MOF/MWCNTs with mapping mode were exhibited in Table S4. Figure S12 and Table S4 illustrate that C-MOF/MWCNTs were mainly composed of C, O, Co elements, and also contained a very small amount of Cu element, and basically did not contain N element. The result was consistent with that of XPS, but the specific content of the elements was different. This may be due to the different accuracy of the test equipment or different sampling sites. The microstructure of C-MOF/MWCNTs was detected via field emission scanning electron microscope (FSME), and the corresponding images at various magnifications were presented in Figure 3. Compared with FSEM images of MOF/MWCNTs ( Figure S3), C-MOF/MWCNTs maintained the fundamental frame structure of MOF/MWCNTs. In detail, MWCNTs still existed after calcination, and calcined MOF retained the normal dodecahedron structure, whereas the surface of MOF altered from firm to porous because of the decomposition of the organic linker during calcination [26]. FSEM image of C-MOF/MWCNTs with mapping mode and its specific energy dispersive X-ray spectroscopy (EDXS) elemental mapping were shown in Figure S12. The mass ratio and atom ratio of elements of FSEM image of C-MOF/MWCNTs with mapping mode were exhibited in Table S4. Figure S12 and Table S4 illustrate that C-MOF/MWCNTs were mainly composed of C, O, Co elements, and also contained a very small amount of Cu element, and basically did not contain N element. The result was consistent with that of XPS, but the specific content of the elements was different. This may be due to the different accuracy of the test equipment or different sampling sites.
Transmission electron microscopy (TEM) images, selected area electron diffraction (SAED) pattern and high-resolution TEM (HRTEM) image of C-MOF were presented in Figure 4. From Figure 4a,b, the surface of C-MOF possessed many holes and was wrapped in MWCNTs, which was in accordance with the FSEM results of C-MOF. Figure 4c shows that SAED pattern of C-MOF revealed that it had the polycrystalline properties. In addition, the lattice planes of Co 3 O 4 ((111), (311)) were observed in the HRTEM image of C-MOF ( Figure 4d). Unfortunately, the lattice planes of CuO were detected via the SAED pattern, HRTEM, and XRD, which were presumably attributed to a low fraction of CuO or amorphous CuO as mentioned above. TEM image of C-MOF with mapping mode ( Figure S13a), EDXS elemental mapping (Figure S13b-f), EDXS curve ( Figure S13g) of C-MOF, and the mass ratio and atom ratio of elements of TEM image of C-MOF with mapping mode (Table S5) also proved, although Cu element's atomic percentage increased slightly, but it was still low.  Transmission electron microscopy (TEM) images, selected area electron diffraction (SAED) pattern and high-resolution TEM (HRTEM) image of C-MOF were presented in Figure 4. From Figure 4a,b, the surface of C-MOF possessed many holes and was wrapped in MWCNTs, which was in accordance with the FSEM results of C-MOF. Figure 4c shows that SAED pattern of C-MOF revealed that it had the polycrystalline properties. In addition, the lattice planes of Co3O4 ((111), (311)) were observed in the HRTEM image of C-MOF ( Figure 4d). Unfortunately, the lattice planes of CuO were detected via the SAED pattern, HRTEM, and XRD, which were presumably attributed to a low fraction of CuO or amorphous CuO as mentioned above. TEM image of C-MOF with mapping mode ( Figure S13a), EDXS elemental mapping (Figure S13b-f), EDXS curve ( Figure S13g) of C-MOF, and the mass ratio and atom ratio of elements of TEM image of C-MOF with mapping mode (Table S5) also proved, although Cu element's atomic percentage increased slightly, but it was still low.  The above characterization revealed that with the decomposition of organic ligands, MOF were converted to major Co3O4 and minor CuO during the calcination at 500 °C for 3 h in air atmosphere.
Thermo gravimetric analyzer (TGA) curves and DTG (derivative thermogravimetry) curves under N2 and air, DTG curve under N2 and air of C-MOF/MWCNTs, respectively, were presented in Figure 5a,b. From Figure 5a,b, there was almost no mass loss for C-MOF/MWCNTs either under N2 or air, because the organic ligands had almost been decomposed during the calcination to prepare C-MOF/MWCNTs. It was worth mentioning that the mass of C-MOF/MWCNTs started to increase slowly after 550 °C under air because of the introduction of oxygen into the residue. Apparently, C-MOF/MWCNTs had better thermal stability compared to MOF/MWCNTs. The above characterization revealed that with the decomposition of organic ligands, MOF were converted to major Co 3 O 4 and minor CuO during the calcination at 500 • C for 3 h in air atmosphere.
Thermo gravimetric analyzer (TGA) curves and DTG (derivative thermogravimetry) curves under N 2 and air, DTG curve under N 2 and air of C-MOF/MWCNTs, respectively, were presented in Figure 5a,b. From Figure 5a,b, there was almost no mass loss for C-MOF/MWCNTs either under N 2 or air, because the organic ligands had almost been decomposed during the calcination to prepare C-MOF/MWCNTs. It was worth mentioning that the mass of C-MOF/MWCNTs started to increase slowly after 550 • C under air because of the introduction of oxygen into the residue. Apparently, C-MOF/MWCNTs had better thermal stability compared to MOF/MWCNTs. MOF were converted to major Co3O4 and minor CuO during the calcination at 500 °C for 3 h in air atmosphere.
Thermo gravimetric analyzer (TGA) curves and DTG (derivative thermogravimetry) curves under N2 and air, DTG curve under N2 and air of C-MOF/MWCNTs, respectively, were presented in Figure 5a,b. From Figure 5a,b, there was almost no mass loss for C-MOF/MWCNTs either under N2 or air, because the organic ligands had almost been decomposed during the calcination to prepare C-MOF/MWCNTs. It was worth mentioning that the mass of C-MOF/MWCNTs started to increase slowly after 550 °C under air because of the introduction of oxygen into the residue. Apparently, C-MOF/MWCNTs had better thermal stability compared to MOF/MWCNTs.

Characterization of RPUFs
RPUFs' chemical component has a tremendous influence on their various properties, such as mechanical properties, flame retardant properties, etc. FT-IR and XPS were carried out identifying the chemical component and structure of RPUFs, and the corresponding

Characterization of RPUFs
RPUFs' chemical component has a tremendous influence on their various properties, such as mechanical properties, flame retardant properties, etc. FT-IR and XPS were carried out identifying the chemical component and structure of RPUFs, and the corresponding testing results were shown in Figures 6 and 7, respectively. As presented in Figure 6, the peaks of all RPUFs at 3331 cm −1 , 2927 cm −1 , 2853 cm −1 , and 1729 cm −1 were respectively assigned to the stretching vibrations of N-H, -CH 3 , -CH 2 -, and C=O, and the peaks at 2274 cm −1 were ascribed to N=C=O anti-symmetric stretching vibrations, and the peaks of N-H bending vibrations were detected at 1593 cm −1 for all RPUFs, and the peaks at 969 cm −1 for P-O stretching vibrations was tested in the FT-IR curves of RPUF-T, RPUF-T/MOF/MWCNTs 2 and RPUF-T/C-MOF/MWCNTs 2 on account of the insertion of TBPBP into the copolyester chains [11,27].  The FT-IR results prove that all RPUFs were successfully prepared, while the peaks of Co and Cu elements were not obviously detected in XPS curves of all RPUFs (Figure 7). The weak peaks of P2p were only found in RPUF-T and RPUF-T/MOF/MWCNTs 2 and  The physical and mechanical properties of RPUFs are profoundly influenced by their cell morphology [28,29]. The cell morphology of all RPUFs were observed via FSEM images (  The FT-IR results prove that all RPUFs were successfully prepared, while the peaks of Co and Cu elements were not obviously detected in XPS curves of all RPUFs (Figure 7). The weak peaks of P2p were only found in RPUF-T and RPUF-T/MOF/MWCNTs 2 and RPUF-T/C-MOF/MWCNTs 2, because TBPBP reacted with pMDI and formed the foam matrix, which was likely to be easier to detect.
The physical and mechanical properties of RPUFs are profoundly influenced by their cell morphology [28,29]. The cell morphology of all RPUFs were observed via FSEM images (Figure 8 The mechanical and physical properties considerably restrict the application of RPUFs [30,31]. RPUFs' mechanical and physical properties generally include density, compressive strength, and thermal conduction coefficient, and the corresponding testing results are summarized in Table 2. According to the test results in Table 2, RPUFs employed MOF/MWCNTs, C-MOF/MWCNTs and TBPBP gained a gradual increase in density, which was considered to be the increase in viscosity of the polyol system [32]. The mechanical and physical properties considerably restrict the application of RPUFs [30,31]. RPUFs' mechanical and physical properties generally include density, compressive strength, and thermal conduction coefficient, and the corresponding testing results are summarized in Table 2. According to the test results in Table 2, RPUFs employed MOF/MWCNTs, C-MOF/MWCNTs and TBPBP gained a gradual increase in density, which was considered to be the increase in viscosity of the polyol system [32]. Fortunately, even the largest density (40.9 kg/m 3 ) of RPUF-T/MOF/MWCNTs 2 was not much greater than the smallest density (36.7 kg/m 3 ) of neat RPUF, which had no influence on the actual application. Compressive strength is an important parameter, which affected the application of RPUFs [33]. Compressive strength of RPUFs is correlated with many factors, such as the formula, foam density, the cellular size, number and its uniformity, and cell wall thickness [34][35][36]. In general, RPUFs with larger density attain larger compressive strength. The  Compressive strength is an important parameter, which affected the application of RPUFs [33]. Compressive strength of RPUFs is correlated with many factors, such as the formula, foam density, the cellular size, number and its uniformity, and cell wall thickness [34][35][36]. In general, RPUFs with larger density attain larger compressive strength. The compressive strength of neat RPUF was 236 ± 29 kPa, while the compressive strengths of RPUF-MOF/MWCNTs 1, RPUF-MOF/MWCNTs 2, RPUF-MOF/MWCNTs 3, RPUF-T, RPUF-T/MOF/MWCNTs 2, RPUF-C-MOF/MWCNTs 2, and RPUF-T/C-MOF/MWCNTs were increased to 301 ± 26 kPa, 358 ± 17 kPa, 379 ± 28 kPa, 422 ± 30 kPa, 453 ± 39 kPa, 367 ± 23 kPa, and 486 ± 25 kPa, respectively, which had an increment of 27.54%, 51.69%, 60.59%, 78.81%, 91.95%, 55.51%, and 105.93%, respectively. The compressive strength was greatly upgraded via using MOF/MWCNTs, C-MOF/MWCNTs, and TBPBP, especially when MOF/MWCNTs and TBPBP, or C-MOF/MWCNTs and TBPBP were used together.
Thermal conduction coefficient is a key index of thermal insulation materials [37]. On the whole, the thermal conduction coefficient of RPUFs was relatively low. Even though the maximum thermal conduction coefficient of RPUF-T/MOF/MWCNTs 2 was 35.53 × 10 −3 W/(mK), which was not much larger than 31.67 × 10 −3 W/(mK) of neat RPUF. The thermal conduction coefficient of RPUFs (λ R ) was calculated by the equation: λ R = λ s + λ g + λ c + λ r , (λ s = thermal conductivity of the solid phase, λ g = thermal conductivity of the gas in the cells, λ c = thermal conductivity attributed to convection within cells, λ r = thermal conductivity assigned to radiation across the cells) [38,39]. As far as the thermal conduction coefficient of other foams was higher than that of neat RPUF for two reasons. On the one hand, the increase in density increased the thermal conduction coefficient for RPUFs, which was assigned to the progressively accrescent contribution of solid thermal conductivity [40]. Additionally, the contribution of solid heat conduction to thermal conduction coefficient is considerably larger than that of thermal convection and thermal radiation [40]. On the other hand, cell size, cell number, and cell wall thickness affected the thermal conduction coefficient of RPUFs via influencing photon transfers, because the thermal conduction coefficient (0.025 mW/(mK) of air is higher than for the CO 2 (0.015 mW/(mK)) [41][42][43].
Thermo gravimetric analyzer (TGA) is commonly used for assessing the thermal stability of polymers [44]. As illustrated in Figure 9, TGA curves and DTG curves of RPUFs were examined under N 2 and air, respectively. Some important indexes, namely the 5% weight loss temperature (T 5% ), the 50% weight loss temperature (T 50% ), the maximum rate of weight loss (Rmax), the temperature of maximum rate (Tmax), and the residue yield, were exhibited in Table S6 (under N 2 ) and Table S7 (under air). In other words, with the increasing amount of MOF/MWCNTs, RPUFs had gradually better thermal stability under air. However, the thermal stability of RPUFs under air was worse than that under N2 because of the presence of oxidizing gases [48]. In terms of T5%, T50%, residue (wt%), and other indicators, RPUF-C-MOF/MWCNTs 2 had better thermal stability than RPUF-MOF/MWCNTs 2. Because the MOF/MWCNTs in RPUF-MOF/MWCNTs 2 continued to decompose during the TGA test, while the C-MOF/MWCNTs in RPUF-C-MOF/MWCNTs 2 was strongly stable at high temperature as mentioned earlier, resulting in the higher thermal stability of RPUF-T/C-MOF/MWCNTs 2 than that of RPUF-T/MOF/MWCNTs 2. Vertical burning test and limiting oxygen index (LOI) test are simple and effective methods to investigate the flame retardancy of RPUFs [28,50]. The flame retardant performance of the vertical burning test of RPUFs is assessed via UL-94 ranks. UL-94 ranks and LOI values of RPUFs were listed in Table 3. From Table 3, neat RPUF failed to obtain any UL-94 rank on account of the low break-up temperature of urethane bonds and high porous cellular structure that caused its rapid combustion [46,51]. RPUF-MOF/MWCNTs 1, RPUF-MOF/MWCNTs 2, RPUF-MOF/MWCNTs 3, and RPUF-C-MOF/MWCNTs 2 all By observing Figure 9, two main degradation processes existed for all RPUFs under both N 2 and air atmospheres; the first degradation stage was assigned to the dissociation of the low-molecular-weight compounds, as polyurethane hard segments, and the second degradation stage was considered to the dissociation of the high-molecular-weight compounds, as polyurethane soft segments [45][46][47]. It is interesting that the first degradation stage under both N 2 and air atmospheres had the same pyrolysis mechanism and all happened ca.200-350 • C, while the temperature of the second degradation stage under air (400-600 • C) was higher than that under N 2 (350-500 • C) and the degradation rate under air accelerated significantly due to the oxidation reaction caused by oxygen [48]. Thus, the residue mass of the same sample under air was much lower than that under N 2 . As for the influences of MOF/MWCNTs or C-MOF/MWCNTs on the thermal degradation of RPUFs under N 2 , T 5% and T 50% were both increased and Rmax 1 was decreased; while Rmax 2 was not changed abruptly, the residue at 700 • C was significantly increased. All in all, the utilization of MOF/MWCNTs or C-MOF/MWCNTs independently improved the thermal stability of RPUFs. Remarkably, when TBPBP and MOF/MWCNTs or TBPBP and C-MOF/MWCNTs were used together, the prepared RPUF-T/MOF/MWCNTs 2 or RPUF-T/C-MOF/MWCNTs 2 obtained more residues in all RPUFs amount (25.79 wt% or 27.39 wt%), revealing that they possessed good thermal stability. T 5% of RPUF-T, RPUF-T/MOF/MWCNTs 2, and RPUF-T/C-MOF/MWCNTs 2 were both ca. 200 • C, lower than these of other RPUFs (over 250 • C), because there was the preferential decomposition of TBPBP due to the early decomposition and poor thermal stability of P-O-C and P-C bonds compared to the C-C bond [11,49]. It was precisely because the preferential decomposition of TBPBP and the simultaneous use of MOF/MWCNTs or C-MOF/MWCNTs promoted the formation of expanded and stable char residues; these were the reasons why RPUF-T/MOF/MWCNTs 2 and RPUF-T/C-MOF/MWCNTs 2 obtained the good thermal stability.
The effect of MOF/MWCNTs or C-MOF/MWCNTs on the thermal degradation of RPUFs under air was similar to those under N 2 . In other words, with the increasing amount of MOF/MWCNTs, RPUFs had gradually better thermal stability under air. However, the thermal stability of RPUFs under air was worse than that under N 2 because of the presence of oxidizing gases [48]. In terms of T 5% , T 50% , residue (wt%), and other indicators, RPUF-C-MOF/MWCNTs 2 had better thermal stability than RPUF-MOF/MWCNTs 2. Because the MOF/MWCNTs in RPUF-MOF/MWCNTs 2 continued to decompose during the TGA test, while the C-MOF/MWCNTs in RPUF-C-MOF/MWCNTs 2 was strongly stable at high temperature as mentioned earlier, resulting in the higher thermal stability of RPUF-T/C-MOF/MWCNTs 2 than that of RPUF-T/MOF/MWCNTs 2.
Vertical burning test and limiting oxygen index (LOI) test are simple and effective methods to investigate the flame retardancy of RPUFs [28,50]. The flame retardant performance of the vertical burning test of RPUFs is assessed via UL-94 ranks. UL-94 ranks and LOI values of RPUFs were listed in Table 3. From Table 3, neat RPUF failed to obtain any UL-94 rank on account of the low break-up temperature of urethane bonds and high porous cellular structure that caused its rapid combustion [46,51]. RPUF-MOF/MWCNTs 1, RPUF-MOF/MWCNTs 2, RPUF-MOF/MWCNTs 3, and RPUF-C-MOF/MWCNTs 2 all did not pass any UL-94 ranks and fast burned in air as well as neat RPUF. this was because MOF/MWCNTs or C-MOF/MWCNTs mixed in RPUFs, which were unable to promote forming the rigid, stable, sufficient chars or releasing non-flammable gases prevented the exchange of heat and combustible gases during RPUFs' combustion. RPUF-T, RPUF-T/MOF/MWCNTs 2 and RPUF-T/C-MOF/MWCNTs 2 all achieved V-0 rank of UL-94 standards, because TBPBP expedited forming the intact, firm phosphorous-nitrogencontaining chars and releasing non-flammable gases containing P or N elements [11]. LOI value is the lowest oxygen content to maintain combustion. It is clear that high LOI values mean good flame retardant performance for RPUFs. As showed in Table 3 FSEM images of RPUFs' char residues obtained via ca. 1300 • C flame were presented in Figure 10. The char residue of neat RPUF was loose, holey, and brittle, which could not protect the interior matrix to evade combustion, and the char residues of RPUF-T were in-tumescent and intact, which could protect the interior matrix to avoid the exchange with heat and oxygen and did not play the role of flame retardant performance. The complete, firm char residues of RPUF-MOF/MWCNTs 1, RPUF-MOF/MWCNTs 2, RPUF-MOF/MWCNTs 3, and RPUF-C-MOF/MWCNTs 2 were both found via FSEM on a microscopic scale, but the char residues were formed on a small scale in the later stages of RPUFs combustion and failed to play a good flame retardant effect. The char residues of RPUF-T, RPUF-T/MOF/MWCNTs 2, RPUF-T/C-MOF/MWCNTs 2 were firm, intact, and intumescent, which acted as flame-retardant barriers in the condensed phase, which resulted in more foam matrix turning into char residues instead of flue gas that pollutes the environment. Cone calorimeter (CC) test can simulate real combustion of RPUFs and acquire sufficient combustion data including the time to ignition (TTI), average effective heat combustion (av-EHC), heat release rate (HRR), total heat release rate (THR), peak heat release rate (pHRR), smoke production rate (SPR), total smoke production (TSP), peak smoke production rate (pSPR), CO production (COP), total COP (TCOP), CO2 production (CO2P), total CO2P (TCO2P), and char residues etc. Combined with the above results and analysis, six RPUFs (namely, neat RPUF, RPUF-MOF/MWCNTs 2, RPUF-T, RPUF-T/MOF/MWCNTs 2, RPUF-C-MOF/MWCNTs 2, and RPUF-T/C-MOF/MWCNTs 2) were chosen for CC test. The corresponding curves of HRR, THR, SPR, TSP, COP, TCOP, CO2P, and TCO2P are presented in Figure 11.
From Figure 11a, the pHRR of RPUFs (except neat RPUF) were all lower than that of neat RPUF, which disclosed that the use of MOF/MWCNTs or C-MOF/MWCNTs or TBPBP decreased the pHRR and enhanced the fire safety during the combustion of RPUFs, especially the simultaneous use of C-MOF/MWCNTs and TBPBP. Notably, all HRR curves were not smooth and divided into several large and small peaks, which was caused by the combustion process of intermediary decomposition products, intermediary char, or volatiles [52][53][54]. THR curves of RPUFs were shown in Figure 11b. The THR values displayed three results. The first result was that both MOF/MWCNTs and C-MOF/MWCNTs slightly improved the fire safety by promoting the formation of a stable carbon layer. Second, TBPBP outstandingly strengthened the fire safety thanks to the formation of the intact, intumescent, sufficient char residues and the release of non-combustible gases (including CO2, NH3, H2O, phosphorus-containing gas, and so on) at high temperature [11]. The third result was that the simultaneous use of MOF/MWCNTs and TBPBP or the simultaneous use of C-MOF/MWCNTs and TBPBP both improved the fire resistance more effectively than the use of MOF/MWCNTs, C-MOF/MWCNTs, or TBPBP alone, due to Cone calorimeter (CC) test can simulate real combustion of RPUFs and acquire sufficient combustion data including the time to ignition (TTI), average effective heat combustion (av-EHC), heat release rate (HRR), total heat release rate (THR), peak heat release rate (pHRR), smoke production rate (SPR), total smoke production (TSP), peak smoke production rate (pSPR), CO production (COP), total COP (TCOP), CO 2 production (CO 2 P), total CO 2 P (TCO 2 P), and char residues etc. Combined with the above results and analysis, six RPUFs (namely, neat RPUF, RPUF-MOF/MWCNTs 2, RPUF-T, RPUF-T/MOF/MWCNTs 2, RPUF-C-MOF/MWCNTs 2, and RPUF-T/C-MOF/MWCNTs 2) were chosen for CC test. The corresponding curves of HRR, THR, SPR, TSP, COP, TCOP, CO 2 P, and TCO 2 P are presented in Figure 11.
From Figure 11a, the pHRR of RPUFs (except neat RPUF) were all lower than that of neat RPUF, which disclosed that the use of MOF/MWCNTs or C-MOF/MWCNTs or TBPBP decreased the pHRR and enhanced the fire safety during the combustion of RPUFs, especially the simultaneous use of C-MOF/MWCNTs and TBPBP. Notably, all HRR curves were not smooth and divided into several large and small peaks, which was caused by the combustion process of intermediary decomposition products, intermediary char, or volatiles [52][53][54]. THR curves of RPUFs were shown in Figure 11b. The THR values displayed three results. The first result was that both MOF/MWCNTs and C-MOF/MWCNTs slightly improved the fire safety by promoting the formation of a stable carbon layer. Second, TBPBP outstandingly strengthened the fire safety thanks to the formation of the intact, intumescent, sufficient char residues and the release of non-combustible gases (including CO 2 , NH 3 , H 2 O, phosphorus-containing gas, and so on) at high temperature [11]. The third result was that the simultaneous use of MOF/MWCNTs and TBPBP or the simultaneous use of C-MOF/MWCNTs and TBPBP both improved the fire resistance more effectively than the use of MOF/MWCNTs, C-MOF/MWCNTs, or TBPBP alone, due to their synergistic fire resistance. The specific synergistic fire resistance was that TBPBP released non-flammable gas and formed an expanded, complete, and stable carbon residue layer with C-MOF/MWCNTs or MOF/MWCNTs (seen Figure 12f and the residues in Table 4 below) because of the increase in viscosity of the carbon residue system, thereby diluting the oxygen concentration and preventing the transfer of oxygen and heat to the interior of RPUF. Meanwhile, the simultaneous use of C-MOF/MWCNTs and TBPBP was better than the simultaneous use of MOF/MWCNTs and TBPBP to improve the fire resistance of foams, because the MOF/MWCNTs continued to decompose in a fire (seen Figure S7), and the resulting char residues are less stable than that of the simultaneous use of C-MOF/MWCNTs and TBPBP (seen Figure S8).  Table 4.

Mechanism for Enhancing Fire and Smoke Safety
Among all RPUFs, RPUF-T/C-MOF/MWCNTs 2 achieved the best flame retardancy and smoke suppression owing to the synergistic effect of C-MOF/MWCNTs and TBPBP. Specifically, TBPBP released non-flammable gas at high temperature, including CO2, NH3, H2O, phosphorus-containing gas, and so on [11]; meanwhile, using TBPBP and C-MOF/MWCNTs promoted the formation of a firm, complete, intumescent, maximum amount of carbon residue layer with C-MOF/MWCNTs because of the increase in viscosity of carbon residue system, thereby diluting the oxygen concentration and  In general, smoke is a crucial factor for influencing the evacuation in the fire accident and polluting the environment [55]. Smoke release during the combustion of RPUFs was recorded via SPR and TSP curves, which were presented in Figure 11c It is internationally conceded that the toxic CO comes from incomplete combustion of materials, and the catalytic conversion of CO to non-toxic CO 2 is a common method to reduce CO concentration. COP curves and TCOP curves were shown in Figure 11e,f, respectively. From Figure 11e, the peak of COP of RPUF-T quickly achieved a fearsome 0.0126 g/s, which was because the formation of the intumescent and intact char residues that were promoted by the early decomposition of TBPBP obstructed the sufficient burning of the foam matrix. Meanwhile, the peak of COP of RPUF-T/MOF/MWCNTs 2 decreased to 0.0097 g/s, probably due to the conversion of CO to CO 2 by the catalysis of CO 3 O 4 and CuO derived from the thermal decomposition of MOF/MWCNTs in a fire. Notably, the peak of COP of RPUF-T/C-MOF/MWCNTs 2 only reached 0.0056 g/s because of the conversion of CO to CO 2 by the catalysis of CO 3 O 4 and CuO in C-MOF/MECNTs. Meanwhile, TCOP of RPUF-T/C-MOF/MWCNTs 2 (0.34 g) were remarkably reduced compared to these of RPUF-T (0.54 g) and RPUF-T/MOF/MWCNTs 2 (0.46 g). The above results disclosed that the introduction of C-MOF/MWCNTs 2 into RPUF obviously decreased the COP during the combustion of RPUF, effectively facilitating people's escape. CO 2 P curves ( Figure 11g) and TCO 2 P curves (Figure 11h) reveal that compared with neat RPUF and RPUF-T, the use of MOF/MWCNTs or C-MOF/MWCNTs did not increase CO2P. The phenomenon could be caused by two reasons. On the one hand, although CO was catalyzed to CO 2 , the CO release was much lower than the CO 2 release during the combustion of RPUFs. On the other hand, the employment of MOF/MWCNTs or C-MOF/MWCNTs improved the char residue stability of RPUFs, resulting in facilitating the conversion of some foam matrix to the char residues instead of CO 2 , which was confirmed by the digital photos of char residues after CC test of RPUFs in Figure 12 and the amount of char residues in Table 4. From Figure 12, compared to the char residue of RPUF-T, both RPUF-T/MOF/MWCNTs 2 and RPUF-T/C-MOF/MWCNTs 2 possessed the firmer, more holistic, and intumescent char residues because of the synergistic effect of MOF/MWCNTs and TBPBP or C-MOF/MWCNTs and TBPBP, which were consistent with the amount of char residues of RPUF-T (9.29 wt%), RPUF-T/MOF/MWCNTs 2 (11.83 wt%), and RPUF-T/C-MOF/MWCNTs 2 (14.18 wt%) in Table 4.
TTI value reveals whether the material is easily ignited. The larger TTI value means the better ignition resistance for materials. Both MOF/MWCNTs and C-MOF/MWCNTs failed to increase TTI values for RPUFs, agreeing with low LOI values and UL-94 ranks. While RPUF-T/MOF/MWCNTs 2 and RPUF-T/C-MOF/MWCNTs 2 had high TTI values like RPUF-T owing to the char-promoting effect of TBPBP, the synergistic effect of TBPBP and MOF/MWCNTs or C-MOF/MWCNTs. Av-EHC displays the degree of burning of volatile gases produced from the materials during combustion [56,57]. Av-EHC values of RPUF-MOF/MWCNTs 2 (20.23 MJ/kg) and RPUF-C-MOF/MWCNTs 2 (22.04 MJ/kg) were not changed much in contrast with that of neat RPUF (21.36 MJ/kg), disclosing that using MOF/MWCNTs or C-MOF/MWCNTs alone had a little effect on Av-EHC for RPUFs. Av-EHC values of RPUF-T, RPUF-T/MOF/MWCNTs 2 and RPUF-T/C-MOF/MWCNTs 2 decreased to 15.75 MJ/kg, 14.97 MJ/kg, and 13.26 MJ/kg, respectively, suggesting that RPUFs inserting TBPBP decomposed phosphorus-containing inhibitors to interrupt the combustion of RPUFs and generate the decrease of Av-EHC [11].

Mechanism for Enhancing Fire and Smoke Safety
Among all RPUFs, RPUF-T/C-MOF/MWCNTs 2 achieved the best flame retardancy and smoke suppression owing to the synergistic effect of C-MOF/MWCNTs and TBPBP. Specifically, TBPBP released non-flammable gas at high temperature, including CO 2 , NH 3 , H 2 O, phosphorus-containing gas, and so on [11]; meanwhile, using TBPBP and C-MOF/MWCNTs promoted the formation of a firm, complete, intumescent, maximum amount of carbon residue layer with C-MOF/MWCNTs because of the increase in viscosity of carbon residue system, thereby diluting the oxygen concentration and preventing the transfer of oxygen and heat into the interior of RPUF. Thus, RPUF-T/C-MOF/MWCNTs 2 had the binary flame retardant mode of condensed phase flame retardant and gas phase flame retardant. Actually, RPUF-T/C-MOF/MWCNTs 2 gained an outstanding effect on reducing the amount of CO release, because CO 3 O 4 and CuO catalyzed the conversion of CO to CO 2 [58]. When CO 3 O 4 and CuO catalyzed the conversion of CO into CO 2 , CO 3 O 4 was converted into CoO, and CuO was converted into Cu. Under the action of high temperature and oxygen, CoO and Cu were again converted into CO 3 O 4 and CuO for participating in the catalysis of CO. A possible mechanism for enhancing fire and smoke safety of RPUF-T/C-MOF/MWCNTs 2 are respectively shown in Scheme 3. MOF/MWCNTs 2 had the binary flame retardant mode of condensed phase flame retardant and gas phase flame retardant. Actually, RPUF-T/C-MOF/MWCNTs 2 gained an outstanding effect on reducing the amount of CO release, because CO3O4 and CuO catalyzed the conversion of CO to CO2 [58]. When CO3O4 and CuO catalyzed the conversion of CO into CO2, CO3O4 was converted into CoO, and CuO was converted into Cu. Under the action of high temperature and oxygen, CoO and Cu were again converted into CO3O4 and CuO for participating in the catalysis of CO. A possible mechanism for enhancing fire and smoke safety of RPUF-T/C-MOF/MWCNTs 2 are respectively shown in Scheme 3.

Conclusions
Compared to neat RPUF, RPUF-MOF/MWCNTs 2 and RPUF-C-MOF/MWCNTs 2 had a certain improvement in physical properties such as the compressive performance, thermal stability, etc., disclosing MOF/MWCNTs and C-MOF/MWCNTs all enhanced physical properties of RPUFs. However, RPUF-MOF/MWCNTs 2 and RPUF-C-MOF/MWCNTs 2 did not achieve good flame retardant properties, because none of them formed firm char residue layers, or generated inhibitory gases that were conducive to flame retardancy. Meanwhile, RPUF-MOF/MWCNTs 2 made no progress in reducing the release of COP in contrast to neat RPUF on account of less Co3O4, CuO created via MOF in MOF/MWCNTs during combustion. However, compared to RPUF-MOF/MWCNTs 2, RPUF-C-MOF/MWCNTs 2 possessed more Co3O4, CuO, resulting in the better improvement in suppressing the emission of THR, TSP, COP, and CO2P, especially the 55.17% COP drop compared to neat RPUF. RPUF-T/C-MOF/MWCNTs 2 gained the best flame retardancy, smoke suppression, CO emission reduction, as well as the best physical properties (including compressive properties, thermal stability, etc.) among all RPUFs because of the synergistic effect of TBPBP and C-MOF/MECNTs, providing a certain reference significance for fabricating RPUFs with human safety and environmental protection as building insulation materials.

Conclusions
Compared to neat RPUF, RPUF-MOF/MWCNTs 2 and RPUF-C-MOF/MWCNTs 2 had a certain improvement in physical properties such as the compressive performance, thermal stability, etc., disclosing MOF/MWCNTs and C-MOF/MWCNTs all enhanced physical properties of RPUFs. However, RPUF-MOF/MWCNTs 2 and RPUF-C-MOF/MWCNTs 2 did not achieve good flame retardant properties, because none of them formed firm char residue layers, or generated inhibitory gases that were conducive to flame retardancy. Meanwhile, RPUF-MOF/MWCNTs 2 made no progress in reducing the release of COP in contrast to neat RPUF on account of less Co 3 O 4 , CuO created via MOF in MOF/MWCNTs during combustion. However, compared to RPUF-MOF/MWCNTs 2, RPUF-C-MOF/MWCNTs 2 possessed more Co 3 O 4 , CuO, resulting in the better improvement in suppressing the emission of THR, TSP, COP, and CO 2 P, especially the 55.17% COP drop compared to neat RPUF. RPUF-T/C-MOF/MWCNTs 2 gained the best flame retardancy, smoke suppression, CO emission reduction, as well as the best physical properties (including compressive properties, thermal stability, etc.) among all RPUFs because of the synergistic effect of TBPBP and C-MOF/MECNTs, providing a certain reference significance for fabricating RPUFs with human safety and environmental protection as building insulation materials.  Table S1: The mass ratio and atom ratio of elements of FSEM image of MOF/MWCNTs with mapping mode; Table S2: The mass ratio and atom ratio of elements of TEM image of MOF with mapping mode; Table S3: The proportion of elements of XPS curves of MOF/MWCNTs and C-MOF/MWCNTs; Table S4: The mass ratio and atom ratio of elements of FSEM image of C-MOF/MWCNTs with mapping mode; Table S5: The mass ratio and atom ratio of elements of TEM image of C-MOF with mapping mode; Table S6: TGA and DTG data of RPUFs under N2; Table S7: TGA and DTG data of RPUFs under air; Table S8