Comparative Study on the Flame Retardancy and Retarding Mechanism of Rare Earth (La, Ce, and Y)-Based Organic Frameworks on Epoxy Resin.

In this work, a series of rare earth-based metal-organic frameworks (RE-MOFs) with the same organic ligand were synthesized and studied as flame retardants on epoxy. Through thermogravimetric analysis, limiting oxide index, UL-94, and cone calorimeter tests, a Y-based MOF (Y-MOF) showed the best flame retardancy compared with a La-based MOF (La-MOF) and Ce-based MOF (Ce-MOF). Further research with Raman, X-ray photoelectron spectroscopy, and theoretical calculation revealed that the reasons for the different flame retardance performances of RE-MOFs resulted from the catalytic carbonizing abilities and the radical-trapping abilities of La, Ce, and Y.

Introduction

Metal–organic frameworks (MOFs) are attractive crystalline porous materials that are constructed by inorganic polynuclear ions/clusters and multidentate organic linkers via coordination bonds. Due to the diverse structure, chemical tunability, and extremely high porosity, MOFs have attracted extensive interest in the past decade.[1−3] Recently, by the sheer breadth of features obtainable in the MOF field, new promises emerged for the development of novel flame retardants (FRs).[4−6] Benefiting from the organic segment of MOFs, MOF-derived FRs showed high compatibility with the polymer matrix and enhanced mechanical property of the polymer composites. Moreover, the nanoscale metal ions/clusters in MOFs play an important role in improving the FR and smoke suppression performances. Up to now, several metal centers have been selected to research the MOF-derived FRs, such as Co,[7−9] Zn,[10−12] Zr,[13−15] Al,[16,17] Fe,[18−20] Cu,[21,22] Ni,[23−25] Sn,[26] and Mo.[27] Hu et al. have found that 2D layered and 3D block MOFs combining common transition metals (such as Fe, Co, Zr, and Ni) and rigid organic complexes showed better flame retardancy with less dosage.[7−9,13,21,25] Zheng et al. reported that Co-, Zn-, and Fe-MOFs can effectively reduce the heat release rate (HRR), smoke production rate (SPR), and CO production (COP) of epoxy (EP) composites.[28] As can be seen, MOFs are potential FRs to obtain superior flame retardancy and fire safety. However, most of these MOFs have different organic ligands or crystal structures; there was no systematic study focus on the metal effect upon the flame retardancy performance. The puzzle of the metal effect will be an impediment for designing MOF-derived FRs. Thus, it is highly desirable to explore the differences of the different metal centers with similar MOF structures as FRs.

Rare-earth (RE) elements share similar chemical properties[29] and many of the RE compounds exhibit good flame retardance properties.[30−38] RE oxides or phosphates, such as lanthanum oxide,[30−34] cerium oxide/cerium phosphate,[35−37] samarium oxide,[39] and so forth, have been widely used as an effective FR synergist and smoke suppressant with the intumescent FRs in polymers. Moreover, layered double hydroxide containing Y[38] or Ce[40] elements not only facilitated the formation of the char layer achieving excellent flame retardancy and smoke suppression but also maintained favorable mechanical properties of the polymer matrix. However, there was no systematic research about RE-MOFs as FRs.

In this work, inspired by the reported flame retardancy of MOFs and RE elements, three RE-MOFs were synthesized with La, Ce, and Y as metal centers and 1,3,5-benzoic acid (BTC) as an organic ligand. The thermal decomposition and flame retardancy of RE-MOFs in the EP matrix were studied; then, an in-depth analysis was carried out to reveal the causes for the different flame retardance performances of RE-MOFs, and a flame-retarding mechanism was proposed. Also, the corresponding mechanical properties of RE-MOF/EP composites were researched.

Experimental Section

Materials

Lanthanum nitrate hexahydrate (La(NO3)3·6H2O), cerium nitrate hexahydrate (Ce(NO3)3·6H2O), and yttrium nitrate hexahydrate (Y(NO3)3·6H2O) were provided by Baotou Research Institute of Rare Earths. EP (E-44, epoxy equivalent = 0.44 mol/100 g) was bought from Nantong Xingchen Synthetic Material Co., Ltd. 1,4-Benzene dicarboxylic acid (BDC), 1,3,5-benzoic acid (BTC), N,N-dimethylformamide (DMF), 4,4′-diaminodiphenylmethane (DDM), and ethanol were purchased from Aladdin Chemical Reagent Co., Ltd. (Shanghai, China).

Preparation of RE-MOFs

La-BTC

In a typical process,[41] 2.165 g (5 mmol) of La(NO3)3·6H2O was placed in 120 mL of deionized water and 1.051 g (5 mmol) of BTC was placed in 120 mL of water/ethanol mixture (v/v = 1:1). Then, the abovementioned two solutions were mixed at 25 °C and stirred using a magnetic stirrer for 1.5 h. Finally, the solid powder product was washed with a water/ethanol mixture (v/v = 1:1) and dried in a vacuum oven at 70 °C.

Ce-BTC

The preparation process of Ce-BTC was similar to La-BTC,[42] and only, 2.165 g (5 mmol) of La(NO3)3·6H2O was replaced by 2.171 g (5 mmol) of Ce(NO3)3·H2O and the time of stirring was reduced to 10 min.

Y-BTC

As a typical preparation procedure,[43] 1.915 g (5 mmol) of Y(NO3)3·6H2O and 1.051 g (5 mmol) of BTC were dissolved in 120 mL of DMF/water mixture (v/v = 5:1), followed by magnetic stirring until the mixture was fully dissolved. Then, the mixture was transferred to a conical flask and kept at 100 °C for 12 h. Finally, the product was washed with water and ethanol, respectively, and dried in a vacuum oven at 70 °C.

Preparation of EP and EP/RE-MOF Composites

EP/RE-MOF composites were prepared using a facile solution blending method. First, 4 wt % as-prepared RE-MOFs were ultrasonically dispersed in a suitable amount of acetone until completely dispersed. Subsequently, the dispersion solution was added into EP and stirred at 90 °C for 2 h. Next, the hardener DDM (ratio of EP/DDM was 10:3) was poured into the abovementioned solution and stirred for 5 min. Finally, the mixture was transferred into the customized Teflon mold, followed by curing at 110, 130, and 150 °C for 2 h, respectively. The pure EP without adding RE-MOFs underwent the same process. Samples with La-BTC, Ce-BTC, and Y-BTC were labeled EP/La-BTC, EP/Ce-BTC, and EP/Y-BTC, respectively.

Characterization

The morphology of the samples was observed on a ZEISS Gemini 300 scanning electron microscope (Germany) at an acceleration voltage of 10 kV. The Brunauer–Emmett–Teller (BET) surface area and porous structures were analyzed based on N2 adsorption–desorption isotherms using an ASAP 2460 (Micromeritics Instrument Corp., USA). X-ray diffraction (XRD) was performed on an Ultima IV instrument (Rigaku, Japan) using Cu Kα radiation (40 kV and 100 mA) in the 2θ range from 10 to 80° with a scan rate of 5° min–1. Thermogravimetric analysis (TGA) was carried out on a TGA 701 instrument (LECO, USA) under N2 conditions with a ramping rate of 10 °C/min. X-ray photoelectron spectroscopy (XPS) was carried out in a Thermo Scientific Escalab 250 Xi spectrometer equipped with a monochromatic Al Kα X-ray source (hv = 1486.6 eV). All XP spectra were recorded using an aperture slot measuring 300 μm × 700 μm. Survey and high-resolution spectra recorded with pass energies (within ±0.2 eV) were determined, with respect to the position of the adventitious C 1s peak at 284.6 eV. Raman spectroscopy was performed with a CRM 2000 laser Raman spectrometer (Renishaw, UK) using a 514 nm laser. The fire behavior was studied using a cone calorimeter test (CCT) (FTT, UK) according to ISO 5660, and the samples (100 × 100 × 4 mm3) were irradiated at a heat flux of 50 kW·m–2. The limiting oxide index (LOI) was tested using a JF-6 Automatic Oxygen Indexer (Jiangning Analytical Instruments, China) according to ASTM D2863, and the sheet dimensions were 100 × 10 × 3 mm3. UL-94 (vertical burning test) was carried on a CZF-5 instrument according to ASTM D3801 with sample dimensions of 130 × 13 × 3 mm3. The tensile properties were tested on a WDW-10M electronic testing machine (Jinan Zhongluchang Instrument, China) according to ISO178 under a rate of 2 mm/min.

Computation

All the first-principles computations were carried out by employing the projector-augmented plane wave[44,45] method as implemented in the Vienna Ab inito Simulation Package (VASP)[46,47] code. The electron exchange–correlation functional was described by the generalized gradient approximation in the form proposed by Perdew, Burke, and Ernzerhof.[48] The tolerance for energy convergence was set to 10–5 eV, while the ionic relaxation is converged when the force on each atom was less than 10–3 eV/Å. The 550 eV cutoff for a plane-wave basis set was adopted in all computations. The Brillouin zone was integrated using Monkhorst–Pack-generated sets of k-points. In the geometry relaxation and self-consistent computations, 7 × 7 × 1 Monkhorst–Pack k-point mesh was used.

Results and Discussion

Characterization of RE-MOFs and EP Composites

RE-MOFs were prepared according to the previous reports, and their crystal structures were confirmed by XRD patterns, as illustrated in Figure S1. For La-BTC and Ce-BTC, both of them have similar characteristic diffraction peaks in the range of 5–50°, which can be well-indexed to a bulk phase of La(BTC) (H2O)6[49] and Ce(BTC)(H2O)6,[42] respectively. With regard to Y-BTC, the strong and narrow diffraction peaks (2θ = 8.6, 10.6, 18.3, and 20.3°) show excellent agreement with the XRD pattern of simulated Y(BTC) (H2O) (DMF)1.1.[43] The morphologies of the RE-MOF crystals were characterized by SEM images, as shown in Figure . It can be seen the shapes of all RE-MOFs are rod-shaped crystals, but compared with La-BTC and Ce-BTC, Y-BTC shows a more regular morphology and larger size. Besides, the porous features of RE-MOFs were studied by N2 physisorption experiments (Figure S2 ad Table S1). The specific surface areas of La-MOF, Ce-MOF, and Y-MOF are 22.18, 16.39, and 2.11 m2/g, respectively, which are very close to the previously reported values (17.3 m2/g for La-MOF,[50−52] 13.0 m2/g for Ce-MOF,[53,54] and 10.0 m2/g for Y-MOF[54,55]). The abovementioned characterizations demonstrate the successful synthesis of RE-MOFs.

Figure 1

SEM images of (a) and (d) La-BTC, (b) and (e) Ce-BTC, and (c) and (f) Y-BTC.

It is well-known that the dispersion degree of FRs in polymer plays an important role in the FR efficiency of the matrix. From the digital photographs of EP and EP composites (Figure a,e,i,m), it can be seen that the EP/RE-MOF samples are quite well-distributed on the macroscale. The microdistributions of RE-MOFs in EP were studied by SEM images of the brittle rupture surfaces. As can be seen, compared with the smooth fracture surface of pure EP (Figure b–d), EP/RE-MOFs showed rather rough and ribbed structural surfaces (Figure f–h,j–l,n–p). The multirib structures may help us to enhance the mechanical properties of the EP matrix. Moreover, most of the RE-MOFs were embedded in EP and there was no obvious aggregation, which indicates that there were well distribution and good interaction between RE-MOFs and the EP matrix.

Figure 2

Digital photographs of (a) EP, (e) EP/La-BTC, (i) EP/Ce-BTC, and (m) EP/Y-BTC and SEM images of brittle rupture surfaces of (b–d) EP, (f–h) EP/La-BTC, (j–l) EP/Ce-BTC, and (n–p) EP/Y-BTC.

Thermal Property and Flame Retardancy

Thermal Properties of RE-MOFs and EP Composites

Thermal stabilities of the obtained RE-MOFs were studied by TGA. From Figure a,b, it can be seen that all the RE-MOFs showed two degradation stages, the first degradation stage (100–300 °C) is corresponding to the volatilization of solvent molecules (H2O or DMF) and the second degradation stage (550–700 °C) is due to the decomposition of the organic groups and MOF structures. Compared with La-BTC and Ce-BTC, Y-BTC exhibited a better thermal stability, which may attribute to its structural integrity upon exposure to heat, vacuum, or pressure treatment.[56]

Figure 3

TG/DTG curves of RE-MOFs (a,b) and EP and EP/RE-MOFs (c,d) under N2 conditions.

The thermal degradation behavior of neat EP and its composites were also measured by TGA, and the corresponding TG/DTG curves are portrayed in Figure c,d. The related data including the initial decomposition temperature (calculated as 5%, T5%) and the char residue are listed in Table . It can be observed that the T5% of EP/RE-MOFs composites is slightly delayed, suggesting that the incorporation of RE-MOFs can give rise to earlier initial degradation of EP composites, which is due to the catalytic decomposition effect of MOFs.[57] The char residue is an important characteristic for the flame retardancy of polymers; generally, the higher the char yield, the better the flame retardancy. TGA results clearly show that the addition of Y-BTC can significantly increase the char yield of EP, while the addition of La-BTC or Ce-BTC has no apparent effect, which means that Y-BTC may have better flame retardancy than La-BTC and Ce-BTC.

Table 1

Data of TGA and DTG of Pure EP and EP Composites

		char yield (%)
samples	T5% (°C)	550 °C	600 °C	650 °C	700 °C
EP	366.1	21.4	18.5	16.9	15.8
EP/La-BTC	360.0	21.5	18.6	16.3	15.4
EP/Ce-BTC	358.1	22.0	20.0	18.4	16.7
EP/Y-BTC	360.7	26.1	24.1	22.9	22.1

Flame Retardancy and Smoke Suppression of EP and Its Composites

The fire safeties of EP composites were first evaluated by LOI measurements and UL-94 vertical burning test. The LOI is the maximum oxygen concentration required to support candle-like combustion in a given testing time and is an important parameter for characterizing the flammability of the polymer materials. As can be seen from Figure , compared with pure EP, the LOI values of EP/RE-MOF composites all increased, especially for EP/Y-BTC, which is up to 29%. Moreover, for EP, the dropping phenomenon occurred and the flammable drops ignited the cotton (Figure b), while the addition of Y-BTC effectively inhibited the melt dropping of EP (Figure c), which can reduce the risk of secondary injury in a fire. These results show that Y-BTC could endow excellent flame retardancy to EP.

Figure 4

(a) LOI and UL-94 data of EP composites and digital photographs of EP (b) and EP/Y-BTC (c) during the UL-94 vertical burning test process.

The combustion behaviors of all samples were further investigated using the CCT, which can simulate bench-scale fire including the release of heat, smoke, suffocating gas (CO and CO2), and so forth. From Figure a, it can be seen that the incorporation of the RE-MOF results in almost no change in the shape of the HRR profiles. However, the highest peak of the HRRs (pHRR) of the EP/RE-MOF composites is reduced in comparison with the pure EP, and the decreasing trend is Y-BTC (28.2%) > Ce-BTC (1.7%) > La-BTC (−2.7%). A similar decreasing trend of the THR values is also observed for the EP/RE-MOF composites, which is Y-BTC (12.7%) > Ce-BTC (−3.2%) > La-BTC (−6.9%) (Figure b and Table ). Furthermore, the char yields of EP/RE-MOF composites have an increasing trend as Y-BTC (120.1%) > Ce-BTC (11.4%) > La-BTC (−15.8%) (Figure c and Table ). To more clearly assess the fire hazards and the performance of flame retardancy of EP/RE-MOF composites, the fire growth index (FGI) and flame retardancy index (FRI) were calculated according to the following equationswhere t-pHRR represents the time to pHRR and TTI represents the time to ignition. The values of t-pHRR and TTI are listed in Table . Generally, with a lower FGI or higher FRI value, EP composites get better fire safety.[58] The value of FGI follows the sequence of EP/Ce-BTC (9.42) > EP/La-BTC (9.01) > EP (8.43) > EP/Y-BTC (6.58), and the value of FRI follows the sequence of EP/La-BTC (0.78) < EP/Ce-BTC (0.92) < EP (1.00) < EP/Y-BTC (1.37). All the abovementioned results indicate that Y-BTC has the most positive effect on the flame retardancy of EP, while La-BTC and Ce-BTC almost have no FR effect on EP.

Figure 5

HRR (a), THR (b), char yield (c) curves, and FGI and FPI values (d) of neat EP and EP/RE-MOF composites.

Table 2

Parameters about Heat, Smoke, and Suffocating Gas in CCT of EP and Its Composites

category	parameter	EP	EP/La-BTC	EP/Ce-BTC	EP/Y-BTC
heat	TTI (s)	43	37	40	37
	pHRR (kW/m2)	1053.3	1081.4	1035.9	756.7
	t-pHRR (s)	125	120	110	115
	THR (MJ/m2)	93.7	100.2	96.7	81.8
	residues (wt %)	11.4	9.6	12.7	25.1
smoke	pSEA (m2/kg)	3918.6	2225.4	3429.9	1434.8
	aSEA (m2/kg)	761.5	741.0	752.8	719.7
	pSPR (m2/s)	0.266	0.290	0.248	0.211
	TSP (m2)	28.2	30.1	28.4	23.9
suffocating gas	pCOP (g/s)	0.030	0.032	0.027	0.021
	pCO2P (g/s)	0.559	0.661	0.598	0.459
	TCO (g)	2.297	2.565	2.192	2.023
	TCO2 (g)	60.80	67.40	64.35	54.50

It is well-known that the released smoke during combustion processes is another significant factor for accessing the fire safety of polymers. From Figure and Table , it is clear that only Y-BTC can effectively reduce the release of the peak of SPR (pSPR) and total smoke production (TSP) at the same time, which is consistent with the results of pHRR and THR. Compared with pure EP, the EP/Y-BTC composite manifests the most effective smoke suppression effect, and the values of pSPR and TSP were reduced by 20.1 and 15.2%, respectively. Moreover, for the EP/Y-BTC composite, another important factor describing the flow rate of combustible gas, the specific extinction area (SEA), also showed a great reduction. For instance, the peak of SEA (pSEA) and the average of SEA (aSEA) were reduced by 63.4 and 5.5%, respectively (Table ). Furthermore, the toxicity gas CO generation of the EP/Y-BTC composite was decreased significantly. As can be seen from Table , the peak of CO production (pCOP) and total CO production (TCOP) were decreased by 30.0 and 11.9%, respectively, in comparison with those of neat EP. However, compared with Y-BTC, La-BTC and Ce-BTC exhibited no or very weak smoke and CO suppression capability.

Figure 6

SPR (a), COP (b), CO2P (c), TSP (d), TCOP (e), and TCO2p (f) curves of neat EP and EP/RE-MOF composites.

Based on these analyses, we can conclude that Y-BTC not only presents a high heat-release suppression performance but also exhibits a strong smoke, CO, and CO2 suppression ability. Moreover, compared with the previously reported MOF FRs (such as Cu-MOF, Co-MOF, Zn-MOF, Fe-MOF, Sn-MOF, Zr-MOF, Al-MOF, Mo-MOF, and Ni-MOF, as shown in Table S2), Y-BTC prepared in this work showed a better comprehensive performance.

FR Mechanism Analysis

Catalytic Carbonizing Abilities of RE Compounds

To analyze the effect of the FR on the char formation, the digital photographs and SEM images of the char residues after the CCT were studied (as shown in Figure ). From Figure a–d, it can be seen that the neat EP almost burns out, resulting in very little char residue with a fragmentary structure, and the EP/La-BTC composite shows loose char residues with a lot of holes on the surface; however, an almost integrated residue was formed both for Ce-BTC and Y-BTC. Furthermore, the SEM images reveal that the residue produced by EP is porous and unconsolidated (Figure e); for EP/La-BTC and EP/Ce-BTC, although there are still a few holes, the residues are relatively integrated (Figure f,g), while for the EP/Y-BTC, the char layer becomes intact and compact, which can act as a protective barrier in suppressing flame transfer (Figure h).

Figure 7

Digital and SEM images of the char residues of the samples after the CCT: EP (a,e), EP/La-BTC (b,f), EP/Ce-BTC (c,g), and EP/Y-BTC (d,h).

The graphitization degree is an important parameter to measure the stability of the char residue, which can be characterized by Raman. As shown in Figure , the D and G bands at 1350 and 1600 cm–1 are corresponding to the peak of amorphous and graphitized carbon, respectively, and the ratio of the D and G band intensity (ID/IG) is used for evaluating the degree of graphitization. It can be seen that the trend of ID/IG is EP/La-BTC (2.46) > EP/Ce-BTC (2.38) > EP (2.14) > EP/Y-BTC (1.58), which indicates that the char residue produced by EP/Y-BTC has the highest degree of graphitization. According to the previous reports about MOF carbonization, MOFs with different metals would form different graphitic degree carbons[27,59−61] and the graphitic degree is determined by the catalytic carbonizing effect of metals. Thus, the high graphitization degree of EP/Y-BTC may be due to the high catalytic carbonizing abilities of Y.

Figure 8

Raman spectra of the char residues of the samples after the CCT: EP (a), EP/La-BTC (b), EP/Ce-BTC (c), and EP/Y-BTC (d).

The chemical structures of EP and its composites were further investigated by XPS analysis to study the thermal oxidative resistance of polymers. It can be found from Figure S3 that the C 1s XPS spectra of all samples could be fitted into three sub-bands at 284.8, 286.0, and 288.8 eV, which are assigned to C–C (aliphatic and aromatic species, represented as Ca), C–O (ether and/or hydroxyl species, represented as Cox), and C=O (carbonyl and/or carboxyl species, represented as Cox), respectively.[27] Generally, the smaller the Cox/Ca ratio, the stronger the thermal oxidation resistance of the char residues. In combination with Figure S2 and Table , it is clear that the Cox/Ca values of EP/RE-MOFs are in the order of EP/La-BTC (0.65) > EP (0.60) > EP/Ce-BTC (0.44) > EP/Y-BTC (0.33), which indicates that Y-BTC would enhance the thermal oxidation stability of the char layer more availably. This result is consistent with that of the Raman analysis.

Table 3

XPS Data of the Char Residues of the Samples after the CCT

	area (%)
samples	C 1s (C–C)	C 1s (C–O)	C 1s (C=O)	Cox/Ca ratio
EP	64.12	30.87	5.01	0.60
EP/La-BTC	60.50	33.67	5.83	0.65
EP/Ce-BTC	69.61	15.31	15.08	0.44
EP/Y-BTC	75.34	11.40	13.26	0.33

Radical Trapping Abilities of RE Compounds

From the abovementioned discussion, we can conclude that compared with La-BTC and Ce-BTC, the excellent FR effect of Y-BTC is benefitted from the high stable carbon formation ability of Y which can prevent the heat transfer and the flame spread. In fact, the thermal degradation of the polymer is a free-radical chain reaction; the radical-trapping ability is also an important factor in reducing the flammability of the polymer.[62] Besides, it was found that the OH• radical plays an important role in the whole thermal oxygen degradation mechanism of the polymer.[63] Thus, the adsorption energy of OH• on the final RE compounds that have direct contact with the flame can be used to evaluate the FR performance. In this work, a computational method was employed to assess the capability of capturing the OH• radical for La, Ce, and Y. According to the XRD analysis of the char residues of the samples after the CCT (Figure S4), the final RE compounds in EP/RE-MOFs were Y2O3, CeO2, and La2CO5, respectively. It should be noted that the La2O3 generated during EP/La-BTC combustion will be transferred to La2CO5 when exposed to atmospheric CO2.[64] Thus, the slab models of Y2O3(222) facets, CeO2(111) facets, and La2CO5(001) facets were constructed to simulate different exposed facets of RE compounds. The OH• group absorbed on the Y2O3(222) facets, CeO2(111) facets, and La2CO5(001) facets were completely optimized to the lowest-energy structures, as shown in Figure .

Figure 9

Optimized geometries of the OH• group absorbed on the (a) Y2O3(222), (b) CeO2(111), and (c) La2CO5(001) facets.

The adsorption energy (Ead) was defined as Ead = ET – (Eslab + Egroup). Accordingly, the smaller adsorption energy indicates the energetically preferable adsorption of free radicals on RE compounds. The calculated adsorption energies of OH• on Y2O3, CeO2, and La2CO5 materials were 0.79, 1.34, and 2.19 eV, respectively. The lowest Ead of the Y2O3 surface implies the highest capability of Y2O3 for capturing the OH• radicals. Additionally, the greatest Ead of OH• adsorbed on La2CO5 suggests the most unfavorability of the adsorption and thus the poorest capability of capturing radicals. As can be seen, the high radical-trapping ability of Y is another important factor for the excellent flame retardance, since it can block the degradation process of EP.

Proposed Mechanism

Based on the abovementioned discussion and analysis, it is interesting to observe that the addition of Y-BTC has the most efficient effect on reducing the heat radiation and suppressing the smoke release, and a possible mechanism is proposed, as shown in Figure . For the pure EP, it is easy to absorb heat and release pyrolysis products without limitation under the action of heating. Large amounts of heat, dense smoke, toxic volatile gases, and suffocating gases (CO and CO2) are released in the case of fire. However, the addition of Y-BTC significantly inhibited the release of a large amount of heat, smoke, and toxic volatile gases, which can be explained by the following reasons. On the one hand, during the pyrolysis process of EP/Y-BTC composites, Y may catalyze the formation of the char layer. The char layer with a superior graphitization degree and better oxidation resistance can be functioned as a barrier to prevent the transfer of heat and decomposition products. On the other hand, the formed Y2O3 can trap the free radicals in the chain reaction and block the degradation process of EP.

Figure 10

Schematic illustration of the proposed mechanism of EP/Y-BTC.

Mechanical Properties of EP and Its Composites

It is well-known that the incorporation of FRs will influence the mechanical properties of the EP matrix. The mechanical performance of EP and its composites was investigated by tensile tests. Figure shows the stress–strain curves of EP and its composites and their mechanical parameters, including tensile strength (σ) and elongation at break (ε). The σ and ε of neat EP can reach about 53.06 MPa and 10.31%, respectively. The addition of La-BTC was slightly averse to the mechanical performance of EP. Different from La-BTC, although the values of EP/Ce-BTC and EP/Y-BTC decreased, their σ values were slightly improved by 7.5 and 1.8%, respectively. In other words, the addition of Ce-BTC and Y-BTC can enhance the EP’s tensile strength but decrease the elongation performance. The tensile strength enhancement may result from the formed multirib structure (as shown in Figure ). The decrease in elongation may be due to the small amounts of pores produced during preparation, which can be removed by improving the preparation method.

Figure 11

Tensile stress–strain curves (a) and tensile strength and elongation at break (b) of EP and its composites.

Conclusions

In this work, three similar RE-MOFs constructed with La, Ce, and Y as metal centers and BTC as an organic ligand were examined as FRs for EP composites. Comprehensive TGA, LOI, UL-94, and CCT results indicated that compared with La-BTC and Ce-BTC, Y-BTC observably enhanced the thermostability and flame retardancy of the EP composites. According to the Raman and XPS analysis of char residues of EP and its composites, the higher catalytic carbonizing effect of Y can substantially improve the graphitization degree and oxidation resistance of the char layer which enhances the flame retardance of Y-BTC. Furthermore, the higher free-radical-trapping ability of Y2O3 may be another contribution to the flame retardance, which can block the chain reaction of EP degradation. Based on the study of this work, the catalytic carbonizing ability and free-radical-trapping ability of the metal center in the MOF should be focused on the design and development of MOF-derived retardants in the future. In addition, the effects of organic ligands and MOF structures on the abovementioned properties will be studied in our next work.