Introduction

Rigid polyurethane foam (RPUF) is widely used in building insulation, refrigeration, transportation, and other fields due to its excellent thermal insulation performance [1]. However, RPUF without flame retardant treatment is easily ignited and will produce toxic and harmful gases such as hydrogen cyanide, nitric oxide, and carbon monoxide [2,3,4], resulting in huge economic losses and personnel injuries. Therefore, the flame retardant modification of RPUF is particularly necessary [5].

Expandable graphite (EG) is an efficient intumescent flame retardant. Its flame-retardant mechanism is that when EG is heated, the acidic substances between the layers are heated to decompose the acidic gas, promote its own expansion and polymer decomposition, and act as a gas source to promote the formation of expanded char layers, thereby blocking the transport of energy and heat [6]. At present, the flame retardant systems for RPUF flame retardant modification include expandable graphite (EG) [7], pentaerythritol phosphate/EG [8], phosphate/EG [9], ammonium polyphosphate (APP)/EG [10], and Yao et al. [11] investigated the synergistic flame retardant effect of APP and EG in semi-rigid polyurethane foam (SPUF). When the flame retardant content was 20% and the mass ratio of EG/APP was 2:1, the LOI value of the composite was increased by about 51% compared with pure SPUF, 2.9% compared with SPUF/EG, and 16.3% compared with SPUF/APP. The SEM results showed that the residual char after combustion of EG/APP/SPUF was dense, and the microstructure was worm-like.

As a typical solid waste, iron tailings (ITS) are mainly composed of alumina, magnesium oxide, iron oxide, and other oxides, and these metal oxides have been successfully used as polymer fillers [12,13,14,15]. Yuan et al. [16] modified γ–Al2O3 on the surface of graphene nanosheets (GO) and introduced it into the polypropylene (PP) matrix to explore the effect of alumina–graphene nanoparticles on PP. The test results showed that the residual char content of PP/γ–Al2O3–GO increased from 0.47–3.85 wt%, and the peak heat release rate (PHRR) decreased by 30.6%. The improvement of flame retardancy was attributed to the barrier effect of graphene and the catalytic char formation of γ–Al2O3 nanosheets.

Based on this, the silane coupling agent was used to modify the surface of ITS, and then, it was compounded with EG to prepare a series of flame retardant rigid polyurethane foam composites in this paper. The thermal stability, combustion characteristics, pyrolysis gas products, and char layer structure of the composites were studied by thermogravimetric (TG), cone calorimeter (CCT), thermogravimetric-infrared (TG-FTIR) and scanning electron microscopy (SEM).

Experimental section

Materials

Polyether polyol (LY-4110; viscosity at 25 °C: 2500 mPa·s; hydroxyl value: 430 mg KOH/g) and triethylenediamine (A33) were bought from Jiangsu Luyuan New Materials Co., Ltd. Polymethylene polyphenyl polyisocyanate (PM200; -NCO content: 30.5–32 wt%; viscosity at 25 °C: 150–250 mPa·s) was purchased from Wanhua Chemical Group Co., Ltd. Silicone oil foam stabilizer (AK8805) was provided by Jining Hengtai Chemical Co., Ltd. Triethanolamine (TEOA) and silane coupling agent (KH550) were purchased by Sinopharm Chemical Reagent Co., Ltd. Catalyst dibutyltin dilaurate (LC) was purchased from Air Products & Chemicals, Inc. Expanded graphite (EG, particle size: 325 mesh, expansion volume: 150 mL/g) was provided by Qingdao Tianhe Graphite Co., Ltd. Iron tailings (ITS; particle size: ≤ 125 um) was kindly supplied by Nanshan Iron Mine in Ma'anshan.

KH550-modified iron tailings

Deionized water (40.0 g) and anhydrous ethanol (160.0 g) were added to a three-neck flask and stirred for 10 min to uniform. Then, the iron tailings (40.0 g) were added to the flask and stirred at 90 °C. After 5 min, KH550 (8.0 g) was added. After 20 min, 5 mL ammonia was added and stirred for 6 h. After filtration and washing, KH550-modified iron tailings (MITS) were obtained by drying at 60 °C.

Preparation of FR-RPUF composites

Table 1 lists the formulation of the RPUF/MITS/EG composites, which were prepared by the one-step water-blown method. According to the ratio in Table 1, reagents other than PM-200 were added to a 500-mL beaker and stirred evenly. Subsequently, PM-200 was added to the mixture and stirred until white and quickly pour into the mold. The sample was cured at 80 °C for 6 h to make the polymerization complete. Finally, the samples were cut into suitable sizes for further characterization.

Table 1 Formulation of RPUF and FR-RPUF composites
Full size table

Measurement and characterization

Fourier infrared (FTIR): ITS or MITS was mixed with potassium bromide (KBr) in a mortar, and then, the tablet was pressed and tested in Nicolet IS5 infrared spectrometer. The test range was 4000–400 cm−1.

X-ray diffraction (XRD) was employed to analyze the mineral composition of char residue of FR-RPUF composites by D8ADVANCE X-ray diffractometer (Bruker Company, Germany). Test condition: Target material: Cu target; Scanning range: −3–150°; accuracy of goniometer: 0.0001°; 2θ angle accuracy: ≤ 0.02°.

The limiting oxygen index (LOI) was tested by the JF-3 oxygen index analyzer (Nanjing Jiangning Instrument Factory, China) in accordance with ASTM D2863 with a sample dimension of 127 mm × 10 mm × 10 mm.

UL-94 vertical burning test was gained using CZF-3 (Nanjing Jiangning Instrument Factory China) in accordance with ASTM-D3801-2010. The sample size was 127 mm × 13 mm × 10 mm.


Thermogravimetric (TG) was conducted on a Q5000IR thermogravimetric analyzer (TA Instruments, USA). 5–10 mg of sample was heated from room temperature to 800 °C with a heating rate of 20 °C/min. Concerning the thermal analyses of RPUF/MITS/EG formulations, the theoretical TG curves were computed. The theoretical TG curve was computed by a linear combination between the TG curves of RPUF/MIT and RPUF/EG.

$$W_{{{\text{th}}}} \left( T \right)_{{\left[ {\text{RPUF/MITS/EG}} \right]}}\, = x \times W_{\exp } \left( T \right)_{{\left[ {\text{RPUF/MITS}} \right]}} + y \times W_{\exp } \left( T \right)_{{\left[ {\text{RPUF/EG}} \right]}}$$
(1)
$$x + y = 1$$
(2)

where \(W_{\exp } \left( T \right)_{{\left[ {\text{RPUF/MITS}} \right]}}\): TG curve of the RPUF/MITS; \(W_{\exp } \left( T \right)_{{\left[ {\text{RPUF/EG}} \right]}}\): TG curve of the RPUF/EG; x, y were the weight percentages of MITS and EG in RPUF, respectively.

Thermogravimetric–Fourier transform infrared spectrometer (TG-FTIR) was performed using Q5000IR (TA Instruments, USA) thermo-analyzer instrument, which linked to Nicolet 6700 FTIR spectrophotometer (Thermo Scientific Nicolet, USA). 5–10 mg of sample was placed in an alumina tray of the TGA and heated from room temperature to 800 °C at 20 °C /min in a nitrogen atmosphere.

Scanning electron microscope (SEM) was performed to obtain the image of RPUF, FR-RPUF, and composites’ char residues by JSM-6490LV scanning electron microscope (JEOL Ltd, Japan). In order to enhance the electrical conductivity of the materials, the samples were coated with a thin conductive layer before observation.

Cone calorimetry was investigated using a conical calorimeter (Fire Testing Technology, UK) in accordance with ISO 5660. The samples with a size of 100 mm × 100 mm × 25 mm were covered with aluminum foil, and the uncovered upper surface was placed horizontally on the sample bench of a cone calorimeter for testing. The thermal radiation flux was set at 35 kW/m2.

The Raman spectra were used to assess the degree of graphitization of residues on an inVia Reflex Laser Raman spectroscopy (Renishaw, UK) with a wavenumber range set from 800–2000 cm−1 and resolution of 1 cm−1.

X-ray photoelectron spectroscopy (XPS) with a VG ESCALAB II spectrometer (Thermo-VG Scientific Ltd USA) using Al Kα excitation radiation (Hν = 1253.6 eV) was used to analyze the char residue of the sample.

Results and discussion

Chemical composition of MITS

Figure 1 shows the infrared spectrum of ITS before and after modification. The peak near 3390 cm−1 was attributed to the stretching vibration of –OH of crystal water in iron tailings [17], and the peak at 1430 cm−1 corresponded to the characteristic absorption peak of \({\text{CO}}_{{3}}^{{2 - }}\) [18]. From the infrared curve of MITS, the absorption peaks at 3620 cm−1 and 2920 cm−1 were attributed to the N–H and –CH2–stretching vibration peaks on aminopropyl, respectively, and the absorption peak at 1020 cm−1 corresponded to the stretching vibration peak of Si–O–Si. The stretching vibration peak of Si–OH at 760 cm−1 [19] and the bending vibration peak of Si–O at 463 cm−1 were both enhanced, indicating that KH550 successfully modified ITS.

Figure 1
figure 1

FTIR spectrum of iron tailings before and after modification

Full size image

Figure 2 shows the thermogravimetric curves of the iron tailings before and after modification, the weight loss of ITS and MITS at 100–400 °C was mainly due to the loss of free water and bound water [20], and the weight loss at high temperature was due to the decomposition of carbonate [21]. At the same time, it could be observed that MITS began to lose weight at low temperature and the mass loss was higher than that of ITS, which was mainly due to the decomposition of KH550 on the surface of MITS [22]. Due to the influence of KH550, the maximum weight loss rate of MIT was significantly lower than that of ITS and the Tmax increased in differential thermogravimetric curve, indicating that KH550 has been successfully coated on the surface of iron tailings.

Figure 2
figure 2

Thermal stability curves of iron tailings before and after modification

Full size image

Mechanical properties

Table 2 provides the density and thermal conductivity of RPUF and FR-RPUF composites. The density of pure RPUF is 64.16 kg/m3. Since the apparent density of EG (about 200 kg/m3) was higher than that of the original RPUF, the incorporation of EG led to an increase in the density of RPUF-6, while the apparent density of FR-RPUF samples partially replaced by MITS decreased, but slightly higher than that of pure samples. Thermal conductivity is an important parameter to characterize the thermal insulation performance of composites. The thermal conductivity of pure RPUF is 0.0366 W/(m·K), indicating that it has good thermal insulation performance. The thermal conductivity of RPUF-6 is 0.0393 W/(m·K), because its density is much higher than that of pure sample. In general, the thermal conductivity of RPUF is affected by density, cell size, open cell number, and foaming agent [23]. With the increase of the proportion of MITS, the thermal conductivity of FR-RPUF decreases and is lower than that of the pure sample, indicating that MITS and/or EG have little effect on the thermal conductivity of the composites, and the composites still maintain excellent thermal insulation performance.

Table 2 Mechanical properties of RPUF and FR-RPUF composites
Full size table

Morphology

The microstructure of RPUF and FR-RPUF composites was observed by SEM to explore the effect of flame retardant particles on the pore structure of polyurethane. As shown in Fig. 3a, the pure RPUF sample had a closed pore structure with a uniform pore size of 300–400 μm [24]. When 20 wt% MITS was added, the pore size of RPUF-2 became larger but the uniformity decreased, and the cell also appeared to have certain damage, which was caused by the poor compatibility between MITS and RPUF matrix. When 20 wt% EG was added, the cell uniformity of RPUF-6 also decreased and some penetrating holes appeared, which was mainly due to the incompatibility between EG particles and the RPUF matrix that would penetrate the cell wall and destroy the cell structure [25]. When 20 wt% MITS/EG (mMITS: mEG = 1:1) was added, the damage to RPUF-4 was more serious, which may be due to the superposition of adverse reactions between MITS, EG and RPUF matrix, which further destroyed the cell structure of the composite.

Figure 3
figure 3

SEM images of RPUF and FR-RPUF composites: a RPUF-1; b RPUF-2; c RPUF-4; d RPUF-6

Full size image

Thermal stability

The effect of MITS/EG on the thermal stability of the RPUF matrix under nitrogen conditions is exhibited in Fig. 4, and the relevant data are listed in Table 3. The results showed that the initial decomposition temperature (T-5wt%) of the pure sample was 273 °C, and the decomposition midpoint temperature (T-50wt%) was 354 °C. The decomposition showed two steps: RPUF hard segment degradation at 250–400 °C and RPUF soft segment degradation at 400–550 °C [26, 27]. The corresponding temperatures (Tmax1 and Tmax2) of the maximum decomposition rates of the two stages were 323 and 470 °C, respectively, and the char residue rate at 700 °C was 19.56 wt%. When 20 wt% MITS was added, the T-5wt% of RPUF-2 decreased to 246 °C, but the T-50wt% increased to 373 °C, and the Tmax1 increased to 335 °C compared with the pure sample, indicating that MITS inhibits the degradation of hard segments of polyurethane. The char residue rate at 700 °C was 29.44 wt%, which was caused by MITS promoting polyurethane carbonization and the non-flammable components in MITS. When 20 wt% EG was added, the T-5wt% of RPUF-6 was almost the same as that of the pure sample, and the T-50wt%, Tmax1, Tmax2, and the char residue rate was increased compared with the pure sample. This was because EG formed an expanded char layer that blocked the transport of oxygen and energy and effectively improved the thermal stability of polyurethane composites. When 20 wt% EG/MITS was added, the data of thermal stability were similar to those of RPUF-6 except that T-50wt% and char residue rate at 700 °C were significantly increased. The char residue rate was 35.76 wt%, which was 82.82% higher than the pure sample, indicating that the combination of EG and MITS effectively improved the high-temperature thermal stability of the composites.

Figure 4
figure 4

TGA a and DTG b curves of RPUF and FR-RPUF composites

Full size image
Table 3 TG data of RPUF and FR-RPUF composites under N2 conditions
Full size table

To further investigate the synergistic effect between MITS and EG, Fig. 5 presents the experimental and theoretical TG curves of RPUF-4. If the calculated curve was a linear combination of the TGA curves of the individual components of the system, it meant that there was no synergistic effect between MITS and EG [28]. As shown in Fig. 5, the experimental curve was always above the theoretical curve, which illustrated that MITS and EG had flame retardant synergistic effect in RPUF composites.

Figure 5
figure 5

The weight difference between the experimental and theoretical TG curves of RPUF/MIT/EG

Full size image

Flame retardant property

The limiting oxygen index (LOI) and vertical combustion (UL-94) were used to evaluate the influence of the MITS/EG system on the flame retardancy of RPUF [29, 30]; the test results are listed in Table 4. The LOI value of RPUF without flame retardant treatment was only 20.1 vol%, and UL-94 was gradeless, so there was a huge potential safety hazard in the use process. When 20 wt% MITS was added, although UL-94 still was gradeless, the molten dripping phenomenon disappeared during the combustion process, and the LOI value increased to 20.7 vol%, indicating that MITS has flame retardancy but the effect was limited. When 20 wt% MITS/EG was added, all RPUF/MITS/EG composites passed UL-94 V-0, and the LOI value also increased significantly (23.3–26.6 vol%). It indicated that MITS and EG had an obvious synergistic effect and effectively reduced the fire risk of RPUF.

Table 4 LOI and UL-94 test results of RPUF and FR-RPUF composites
Full size table

In order to further investigate the effect of MITS and EG on the flame retardancy of RPUF, the combustion behavior of RPUF composites was investigated by a cone calorimeter [31, 32]. The cone calorimeter can provide parameters related to heat release and smoke release, including ignition time (TTI), heat release rate (HRR), total heat release (THR), smoke release rate (SPR), total smoke production (TSP) [33]. Figure 6a presents that the TTI of the pure sample was 4 s, HRR reached its peak (PHRR) which was 442.78 kW/m2 at 51 s after ignition and then drops to zero at about 150 s. When 20 wt% MITS was added, the HRR curve of RPUF-2 was basically coincident with the pure sample, but the PHRR was slightly lower than the pure sample. When 20 wt% EG or MITS/EG was added, the PHRR of the composites decreased significantly, and the PHRR of RPUF-4 and RPUF-6 decreased by 61.55 and 71.99%, respectively, compared with the pure samples. Figure 6b displays that the THR of RPUF-1 and RPUF-2 rose rapidly at a large slope and then rose slowly to the end at a rate close to the horizontal level. THR of the pure sample was 29.7 MJ/m2. RPUF-4 and RPUF-6 rose slowly at a relatively small slope, indicating that the addition of EG or MITS/EG could avoid the heat hazard caused by heat accumulation in a short time. The THR of RPUF-4 and RPUF-6 was 27.3 and 23.76 MJ/m2, respectively.

Figure 6
figure 6

Characteristic curve of RPUF and FR-RPUF composites: a HRR; b THR; c SPR; d TSR; e CO production; f CO2 production; g SF; h mass

Full size image

The smoke release rate (SPR) and total smoke release (TSR) were used to characterize the smoke release of the composites during combustion [34, 35]. Figure 6c presents that the SPR curve of the pure sample gradually increased, reached the peak at 57 s, and then rapidly decreased to zero. When 20 wt% MITS was added, the SPR curve peak of RPUF-2 decreased, and the single peak of RPUF-1 changed into double peaks. When 20 wt% EG or MITS/EG was added, the SPR peaks of RPUF-4 and RPUF-6 decreased sharply by 66.02% compared with the pure sample. The TSR of the pure sample was 630.46 m2/m2, and the TSR of RPUF-2 decreased to 563.47 m2/m2 when 20 wt% MITS was added, indicating that MITS has a certain smoke suppression effect. When 20 wt% EG or MITS/EG was added, the TSR of RPUF-4 and RPUF-6 was significantly reduced, and their values were 138.74 m2/m2 and 66.16 m2/m2, respectively, which were 77.99% and 89.51% lower than those of the pure samples, illustrating that MITS/EG and EG systems have a significant smoke suppression effect. The smoke parameter (SP) is the product of Av-SEA and PHRR. As can be seen from Table 5, the SP of RPUF-4 and RPUF-6 was much lower than that of the pure sample, showing a good smoke suppression effect.

Table 5 Cone calorimetry data of RPUF and FR-RPUF composites
Full size table

The smoke factor (SF) is the product of PHRR and TSR [36]; the fire performance index (FPI) is the ratio of TTI to PHRR. The fire growth rate index (FGI) is the ratio of PHRR to the time to reach the peak (TP), and the three are used to characterize the fire safety of composite materials. The greater the FPI, the smaller the SF and FGI, indicating that the smaller the fire risk of the material. It can be seen from Fig. 6g and Table 5 that when MITS/EG or EG was added, the SF and FGI of the composite decreased significantly, which SF of RPUF-6 was about 91.54% lower than that of the pure sample, and FPI increased compared with pure samples, indicating that MITS/EG effectively reduced the fire risk of composites.

As a typical toxic and harmful gas, CO seriously threatens the safety of people in fire [37]. Figure 6e and f presents that the production rates of CO and CO2 were relatively high after the pure sample was ignited, which was not conducive to the safety of personnel in the fire. Compared with the pure sample, the CO and CO2 production rates of RPUF-4 decreased sharply in the early stage of fire after adding MITS/EG, which provided valuable time for fire escape and rescue and improved the safety of evacuation.

Figure 6 h exhibits the mass change curve during the cone calorimeter test. The mass of the pure sample decreased rapidly after ignition, reaching the final mass (8.33 wt% at about 70 s and the mass-loss rate (MLR) was 4.41 g/s·m−2. After adding 20 wt% MITS, the MLR of RPUF-2 decreased (3.72 g/s·m−2) and the final char residue increased to 11.17 wt%. When 20 wt% MITS/EG was added, the MLR of RPUF-4 was significantly reduced, which was 49.66% lower than that of the pure sample, and the final residual char content was increased to 23.32 wt%, indicating that MITS/EG system could effectively improve the high-temperature stability of the composites.

The flame retardancy index (FRI) is a dimensionless factor that characterizes the flame retardant properties of composites by PHRR, THR and TTI. FRI is defined as the dimensionless ratio between \({\text{THR}}*\left( {\frac{{{\text{PHRR}}}}{{{\text{TTI}}}}} \right)\), where the molecular term corresponds to the neat polymer, and the denominator term corresponds to the composite [38]:

$${\text{FRI}} = \frac{{\left[ {{\text{THR}}*\left( {\frac{{{\text{PHRR}}}}{{{\text{TTI}}}}} \right)} \right]_{{\text{Neat Polymer}}} }}{{\left[ {{\text{THR*}}\left( {\frac{{{\text{PHRR}}}}{{{\text{TTI}}}}} \right)} \right]_{{{\text{Composite}}}} }}$$
(3)

The fame retardancy of composites can be qualitatively expressed using the FRI, with three possible cases, ‘Poor’ (FRI < 1), ‘Good’ (FRI < 10), and ‘Excellent’ (10 < FRI < 102), defined on a logarithmic scale (Table 6). The table indicated that the FRI values of RPUF-2, RPUF-4 and RPUF-6 were greater than 1 after the addition of MITS and/or EG, indicating that MITS and/or EG enhanced the flame retardant properties of FR-RPUF. In addition, the FRI values in some other articles are calculated and summarized in Table 6, from which it could be found that RPUF-4 maintains excellent flame retardant properties.

Table 6 FRI values calculated in this and other articles
Full size table

Figure 7 presents photographs of char residue after the cone calorimeter test of RPUF and FR-RPUF composites. Figure 7 a shows that the surface of the char layer of the pure sample was loose and easy to rupture, which was not conducive to blocking the transport of oxygen and material, so the flame retardancy of pure RPUF was poor. After adding 20 wt% MITS, the char layer of RPUF-2 was damaged more seriously, and the brown inorganic metal oxides were distributed on the surface, indicating that the addition of MITS alone could not promote the formation of a dense char layer of RPUF. When 20 wt% EG or MITS/EG was added, obvious worm-like structures were observed on the surface of char layers of RPUF-4 and RPUF-6, which could play a role in condensed phase flame retardant and effectively improve the fire safety of composites.

Figure 7
figure 7

Photographs of char residue after cone calorimetry of RPUF and FR-RPUF composites: a RPUF-1; b RPUF-2; c RPUF-4; d RPUF-6

Full size image

Gaseous product

TG-FTIR was used to investigate the effect of the MITS/EG system on the release strength of the cracked gas products of the composites [42]. The FTIR spectra of RPUF and FR-RPUF gas products at the maximum degradation rate are displayed in Fig. 8. The infrared peak values of the four materials were basically the same, implying that MITS/EG did not change the degradation mechanism of RPUF composites, and the gas products were still mainly hydrocarbons, isocyanates [43], aromatic compounds, CO2, carbonyl compounds, and esters [44].

Figure 8
figure 8

TG-FTIR spectra of pyrolysis products at a maximal decomposition rate of RPUF and FR-RPUF composites

Full size image

Figure 9 is the typical gas-phase product release strength curve of RPUF and FR-RPUF composites. It can be observed from Fig. 9a that the G-S curves of RPUF-2, RPUF-4 and RPUF-6 were lower than those of pure samples, and the curve of RPUF-4 was lower than that of RPUF-6, indicating that MITS could effectively inhibit the pyrolysis of polyurethane matrix and reduce the formation of its gaseous products. Figure 9b, c, and f shows the release curves of CO2, isocyanates and HCN, respectively. It could be seen that MITS, EG and MITS/EG systems reduce the release intensity of CO2, isocyanates and HCN during the pyrolysis of composites. Figure 9d exhibits the release curve of aromatic compounds. Pure RPUF was a bimodal structure that appears around 900 and 1100 s, respectively. The release intensity of RPUF-2 is higher than that of the pure sample and the peak time of RPUF-2 and RPUF-4 was earlier than that of the pure sample, while the peak time of RPUF-6 was delayed. The peak intensities of RPUF-4 and RPUF-6 were almost the same as those of the pure sample, indicating that MITS/EG system had a certain inhibitory effect on the generation of aromatic compounds. Figure 9e shows the release curves of esters. As shown in the figure, the addition of MITS and EG inhibited the formation of esters, but the MITS/EG system promoted the formation of esters. Overall, the total pyrolysis products of RPUF-2 and RPUF-4 were lower than those of pure samples, indicating that MITS and MITS/EG reduced the release intensity of gaseous products and reduced the release of toxic and combustible gases.

Figure 9
figure 9

Absorbance of pyrolysis products of RPUF and FR-RPUF versus time: a Gram–Schmidt; b CO2; c isocyanate compound; d aromatic compounds; e esters; f HCN

Full size image

Char residue analysis

The morphology of char residue for RPUF and FR-RPUF composite is shown in Fig. 10. It can be seen from Fig. 10a, a1 that the char residue of RPUF was loose, thin, and easy to break. Such a structure was not conducive to preventing the spread of flame and heat and mass transfer, resulting in a low limiting oxygen index. When 20 wt% MITS was added, the thickness of the char layer for RPUF-2 increased slightly, but the pores penetrated through the char layer surface and MITS particles could be obviously observed on the char layer surface. When 20 wt% EG was added, the worm-like structure of EG after combustion expansion can be obviously observed in Fig. 10d, d1 [45], which effectively prevented heat and mass transfer during combustion. Figure 10c, c1 also observes a worm-like structure and the size of the worm-like char layer was larger than RPUF-6, indicating that MITS promoted EG expansion, while MITS mixed with expanded char layer effectively improved the heat resistance of the char layer, thereby protecting the underlying material to achieve the purpose of flame retardant.

Figure 10
figure 10

SEM images of char residues for RPUF and FR-RPUF composites: a: RPUF-1; b: RPUF-2; c: RPUF-4; d: RPUF-6

Full size image

Raman spectra of RPUF and FR-RPUF composites are analyzed to understand the effect of the MITS/EG system on the char formation ability of the RPUF matrix. As shown in Fig. 11, all curves show two characteristic peaks, corresponding to the D peak of amorphous carbon SP3 vibration (1360 cm−1) and the G peak of SP2 vibration with complete graphitization structure (1580 cm−1) [33, 46], respectively. The graphitization degree of the material is expressed by the area ratio of D peak to G peak (ID/IG) [47,48,49]. The smaller ID/IG is, the higher the graphitization degree of the formed char slag is and the better heat resistance of the char slag is [38], so it effectively plays the role of condensed phase flame retardant. The ID/IG of the pure sample was 2.36, while the ID/IG of RPUF-2 increased to 2.48 when 20 wt% MITS was added, which was mainly due to the poor catalytic ability of MITS which was mainly metal oxides alone. When 20 wt% EG was added, the ID/IG of RPUF-6 decreased to 1.96, indicating that EG promoted the formation of the graphitization char layer. When 20 wt% MITS/EG was added, the IiD/IG of RPUF-4 further decreased to 1.85, indicating that MITS/EG system could effectively improve the graphitization degree of the char layer, which was mainly because MITS reacted with EG and RPUF pyrolysis products to form metal ions-carbon compound char layer compounds.

Figure 11
figure 11

Raman spectra of char residue for RPUF and FR-RPUF composites

Full size image

In order to understand the existence of ITS in the char residue, the char residue of RPUF and FR-RPUF composites was analyzed by XRD. As exhibited in Fig. 12, a wide peak appeared in the pure sample at 2θ = 23°, which was caused by amorphous char formed by RPUF combustion [50]. When 20 wt% EG was added, RPUF-6 showed a sharp peak at 2θ = 26.44°, corresponding to the (002) crystal plane of EG [51, 52]. When 20 wt% MITS was added to RPUF-2, some characteristic peaks of ITS disappeared, and the remaining peak area decreased significantly, indicating that there was a certain chemical reaction between MITS and RPUF pyrolysis products. When 20 wt% MITS/EG was added, most of the characteristic peaks of ITS in the XRD curve of RPUF-4 disappeared, and only two obvious peaks at 2θ = 26° and 27° still existed, which were the characteristic peaks of EG and ITS, respectively.

Figure 12
figure 12

XRD spectra of char residue for RPUF and FR-RPUF composites

Full size image

In order to further understand the main elements in the char residue and their bonding state, XPS test was performed on the composites and the results are shown in Fig. 13. It could be seen that the char residue of the RPUF mainly contained C, N and O, and the contents were 67.39, 14.21 and 18.4%, respectively. After adding MITS, the content of C and N in the char residue for RPUF-2 decreased, and the content of O increased. In addition, it also contained 9.92% Si element. Due to the introduction of EG into RPUF-4 and RPUF-6, the C element of the composites increased significantly and the N element decreased.

Figure 13
figure 13

XPS spectra of char residues for RPUF and FR-RPUF composites

Full size image

XPS PEAK41 software was used to perform peak fitting for the main elements C, N and O in the char residues to understand the binding state of the main elements, and the fitting results are shown in Table 7. Figure 14 shows the C (1 s) peak fitting spectra of char residues for RPUF and FR-RPUF composites. The peak at 284.7 eV represented the C element existing in the C–C or C–H bond in the char residues, which corresponded to the graphite carbon structure in the char residues. The peak at 285.5 eV represented C in the form of C–O or C–N bond, which was mainly corresponding to the structures of esters, ethers, alcohols, and phenols on the surface of carbon layer. The peak at 287.4 eV represented C in the form of C = O or C = N, corresponding to the char residue’s surface with carbonyl, carboxylic acid (salt), ester group, aromatic heterocyclic compounds and other structures [53]. The C content in RPUF with medium C–C/C–H, C–O/C–N and C=O/C=N structure was 44.54%, 26.81% and 28.94%, respectively. The addition of MITS resulted in a significant increase in the C–O/C–N content of RPUF-2, and the addition of EG and MITS/EG resulted in a significant increase in the C–C/C–H content, indicating that EG can contribute to a more regular and dense char residue.

Table 7 Bonding state data of C, N, O in char residue of RPUF and FR-RPUF composites
Full size table
Figure 14
figure 14

C1s spectra of char residue for RPUF and FR-RPUF composites: a RPUF-1; b RPUF-2; c RPUF-4; d RPUF-6

Full size image

Figure 15 shows the O (1 s) peak fitting spectra of char residues for RPUF and FR-RPUF composites. The peak at 531.3 eV corresponded to the = O structure of carbonyl, quinone, carboxylic acid (salt) and ester in char residue, the peak at 532.5 eV corresponded to the–O–structure in the form of alcohol, ether and phenol, and the peak at 533.7 eV corresponded to the O element in the form of adsorbed oxygen (O2) and free water (H2O) [54]. The structures of = O, − O − and O2/H2O in RPUF accounted for 49.11, 19.38, and 31.51%, respectively. With the addition of MITS, the O element of RPUF-2 mainly as − O − structure, and its proportion increased to 40.94%. The proportion of O2/H2O structure decreased to 16.08%. When EG was added, the proportion of=O structure in char residue for RPUF-6 increased slightly, and the content of O element in − O − form increased to 35.5%. When MITS/EG was added, the O element of RPUF-4 mainly existed in the form of = O, indicating that the compactness of char residue was significantly improved and MITS/EG changed the bonding state of O element in char residue.

Figure 15
figure 15

O1s spectra of char residue for RPUF and FR-RPUF composites: a RPUF-1; b RPUF-2; c RPUF-4; d RPUF-6

Full size image

Figure 16 shows the N (1 s) peak fitting spectra of char residues for RPUF and FR-RPUF composites. The peak at 398.8 eV corresponded to the N element in the amide and amine compounds with the − NH − structure in the char residues, and the peak at 400.4 eV corresponded to the N in aromatic heterocyclic compounds with = N structure. The MITS and/or EG promoted the development of N element in char residues toward − NH − structure.

Figure 16
figure 16

N1s spectra of char residue for RPUF and FR-RPUF composites: a RPUF-1; b RPUF-2; c RPUF-4; d RPUF-6

Full size image

Figure 17 shows the Si (2p) spectra of char residues for RPUF-2 and RPUF-4 composites. It can be seen from the figure that RPUF-2 had a peak at about 103 eV, indicating the presence of SiO2 in char residue [55], while RPUF-4 had a higher peak intensity, indicating that EG promotes the formation of SiO2.

Figure 17
figure 17

Si2p spectra of char residue for RPUF-2 and RPUF-4 composites

Full size image

Flame retardant mechanism

Based on the above analysis, the flame retardant mechanism of RPUF/MITS/EG composites was proposed. As shown in Fig. 18, the oxide in MIT improved the thermal stability of RPUF composites in the early stage of combustion. When RPUF/MITS/EG composites are heated, acidic substances such as sulfuric acid and nitric acid in EG particles decompose and release a large number of acidic gases during combustion, which promotes the degradation of hard segments of rigid polyurethane foam and the formation of wormlike char layers as gas sources. The metal oxide in MIT played its role by catalytic carbonization effect and can promote the char formation of RPUF composites, and the acidic substances in EG react with metal oxides to form a metal ion carbon layer. The SiO2 produced by the MITS and decomposition of KH550 and the unreacted metal/non-metal oxides are distributed in the worm-like char layer to enhance the condensed phase flame retardant effect. Therefore, MITS/EG mainly plays a condensed phase flame retardant role.

Figure 18
figure 18

Schematic illustration for the flame retardant mechanism of FR-RPUF composites

Full size image

Conclusion

In this paper, modified iron tailings (MITS) were prepared by surface modification of iron tailings (ITS) with KH550, and the synergistic system of MITS/EG flame retardant rigid polyurethane foam (RPUF) was composed of expandable graphite (EG). The effects of a synergistic system on the thermal stability of RPUF matrix and gas-phase products were systematically studied. TG test showed that when 20 wt% EG/MITS was added, the T-50 wt%, Tmax1, and Tmax2 of RPUF-4 were increased to 386, 334, and 486 °C, respectively. The char residue at 700 °C is 35.76 wt%, which is 82.82% higher than that of the pure sample, indicating that the combination of EG and MITS can effectively improve the high-temperature stability of the composites. The CCT test showed that MITS/EG promoted the formation of a dense carbon layer, thereby reducing the thermal and non-thermal hazards of composites. The PHRR, TSR, and SF of RPUF-4 were reduced by 61.55, 77.99, and 91.54%, respectively, compared with the pure sample. TG-FTIR test showed that the release intensity of G–S, CO2, isocyanate compounds, and HCN of RPUF-4 was lower than that of the pure sample after adding MITS/AHP. SEM, Raman and XRD tests show that the synergistic effect of EG and MITS improves the graphitization degree of the carbon slag of the composite.