Flame-retardant functionalized boron nitride for epoxy resin with excellent flame retardancy and enhanced thermal conductivity

Abstract

A flame-retardant functionalized boron nitride (BN@DT) was prepared through the modification of boron nitride (BN) by a reaction product (DT) among 9, 10-dihydro-9-oxa-10-phosphaphenanthrene-10-oxide (DOPO), 4-aminobenzoic acid, and p-phthalaldehyde, and the effect of BN@DT on flame retardancy, thermal conductivity and mechanical properties of epoxy resin (EP) was investigated. The EP composite with 15 wt% BN@DT (EP/15BN@DT) achieves the UL-94 V-0 rating and a limiting oxygen index of 40%. The peak heat release rate, total heat release and total smoke production of EP/15BN@DT are reduced by 36.3%, 31.9% and 38.9%, respectively compared with the ones of pure EP. The thermal conductivity of EP/15BN@DT increases by 148% in comparison with that of pure EP. Moreover, the tensile and impact strengths of EP/15BN@DT are higher than the ones of the EP composite containing 15 wt% unmodified BN. This work provides a promising method for the preparation of the thermally conductive EP composites with excellent flame retardancy.

Keywords

Flame retardancy thermal conductivity epoxy resin functionalized boron nitride

Introduction

In recent years, the miniaturization and integration of electronic and electrical devices have resulted in a continuous increase in power density and new challenges in thermal management.^1,2 The generation of intense heat in microelectronic devices during high-speed power consumption operations not only seriously reduces the working efficiency and service life of the devices, but also may lead to the potential fire hazards.^3,4 Therefore, it is evident that the implementation of thermal management strategies with highly efficient thermally conductive materials represents a crucial factor in ensuring the device life and safety in question.⁵ In this context, preparation of novel thermally conductive materials with multifunctional properties (e.g., flame retardancy and electrical insulation) is of great importance, both for academic research and industrial applications.

Epoxy resin (EP) is extensively used as electronic packaging materials and thermal interface materials due to its excellent mechanical properties, electrical insulation, adhesion characteristics and chemical resistance.^6,7 Nevertheless, the intrinsic low thermal conductivity and high flammability of EP have severely limited its application in advanced electronic products and other key fields.^8,9 Consequently, it is crucial to develop epoxy-based thermally conductive composites with excellent flame retardancy in order to meet the requirements of efficient heat dissipation and fire risk reduction.

In general, the incorporation of fillers into the EP matrix represents the most common method for the fabrication of high-performance thermally conductive EP composites.¹⁰ Thermally conductive fillers generally include metals (e.g. silver and copper nanoparticles), carbon-based materials (e.g. graphene, carbon fibers, carbon nanotubes) and ceramic materials (e.g. boron nitride, silicon dioxide, aluminum oxide).^11,12 Due to their good inherent electrical conductivity, metallic fillers and carbon-based materials are unsuitable for applications where insulation is necessary.¹³ In contrast, the electrical insulation of ceramic-based thermally conductive fillers allows them to be employed in applications that necessitate both high thermal conductivity and high electrical resistivity.¹⁴ Among them, as one of the two-dimensional (2D) layered materials, hexagonal boron nitride (BN) exhibits a large specific surface area similar to graphene and excellent characteristics, including superior electrical insulation, outstanding in-plane thermal conductivity, remarkable thermal stability, and superb lubrication performance.^15,16 Recently, these extraordinary features have stimulated considerable research interests in BN, particularly within the domain of polymer-based thermal management materials, a field of immense importance to the rapidly expanding fifth-generation (5G) communication.^17,18

However, the inert surface and strong interlayer interactions of bulk BN lead to its insufficient dispersion and weak interfacial interactions within the EP matrix, resulting in a considerable thermal resistance at the filler-matrix interface, which substantially obstructs the thermal conductivity enhancement of EP/BN composites.^19,20 Furthermore, EP/BN composites with high thermal conductivity require a high loading of BN, which leads to the formation of agglomeration and stress concentration, ultimately resulting in an adverse impact on the machinability and mechanical properties of the composites.^21,22 To solve the dispersion and compatibility issues of BN, the surface modification (e.g. covalent and non-covalent modification) offers an effective strategy.^23,24 For instance, Ding et al. prepared oxygen-containing groups modified boron nitride nanosheets through a simple ball milling process with the furoic acid as a chemical modifier, which facilitated the dispersion of BN and resulted in the production of aerogels with highly thermal conductivity.²⁵ Jiang et al. found that the modified BN by poly (glycidyl methacrylate) effectively enhanced the compatibility and interfacial bonding strength between BN and the EP matrix, and the thermal conductivity of the EP/modified BN composite reached 1.198 Wm⁻¹K⁻¹, higher than that of the EP/raw BN at the same filler content (15 vol%).²⁶ Therefore, the surface modification of BN represents an effective solution to improve the compatibility between BN and the polymer matrix, as well as reduce the thermal resistance.

On the other hand, given the flammability of EP and the limited flame retardancy of BN in EP, it is also crucial to incorporate an effective flame retardant. Previous research has reported that the functionalization of 2D fillers with specific flame retardants not only facilitated the dispersion of fillers within the polymer matrix and enhanced the thermal conductivity, but also imparted good flame retardancy to the composites.^9,27,28 Phosphorus-containing flame retardants, particularly the 9, 10-dihydro-9-oxa-10-phosphaphenanthrene-10-oxide (DOPO) and its derivatives, are extensively utilized in the flame retardancy of EP composites due to their notable efficiency and low toxicity.²⁹ For example, Yang et al. grafted DOPO onto the surface of BN via covalent modification and found that the functionalized BN improved the thermal conductivity and mechanical properties of EP composites to a greater extent than untreated BN, as well as enhanced the flame retardancy.³⁰ However, the aforementioned EP composites did not pass the UL-94 V-0 rating even at the loading of 20 wt%, which may be attributed to the relatively limited amount of grafted flame retardants on BN. Therefore, non-covalent flame retardant functionalization of BN may be a better choice to ensure the desired flame retardancy.

In this work, a flame-retardant functionalized boron nitride (BN@DT) was prepared by an in-situ method through the modification of boron nitride (BN) by a reaction product (DT) among 4-aminobenzoic acid, p-phthalaldehyde and 9, 10-dihydro-9-oxa-10-phosphaphenanthrene-10-oxide (DOPO), and the effect of BN@DT on the flame retardancy, thermal conductivity, mechanical properties and thermal degradation of EP was studied comprehensively. The aromatic groups in DT enable its immobilization onto the BN surface via π-π interactions, and the polar carboxyl groups form hydrogen bonds with the BN surface to achieve modification during sonication, facilitating the dispersion of BN.^25,31 This work has indicated that the immobilization of DT on the BN surface is critical for enhancing its dispersion and interfacial compatibility, which is also the key to improve the thermal conductivity and mechanical properties of the EP/BN composites. Furthermore, DT exhibits excellent flame-retardant properties in EP, thereby compensating for the limitation of BN’s low flame retardancy.

Experimental

Materials

Ethanol (95%) was provided by Shanghai Boer Chemical Reagent Co., Ltd, China. P-phthalaldehyde (98%), 9,10-dihydro-9-oxa-10-phospha-phenanthrene-10-oxide (DOPO, 97%), 4,4′-diaminodiphenylmethane (DDM, 98%) and 4-aminobenzoic acid (98%) were acquired from Aladdin Industrial Co., Ltd. Boron nitride (BN, 2–8 μm) was obtained from Brofos Nanotechnology (NingBo) Co., Ltd, China. Bisphenol-A epoxy resin (DGEBA, trade name E51, epoxy value 0.51) was purchased from China Sinopec Group Co., Ltd

Synthesis of DT

First, p-phthalaldehyde (2.01 g, 0.015 mol) was dispersed into 100 mL ethanol within a 500 mL three-necked flask equipped with a magnetic stirrer. Then, a solution of 4-aminobenzoic acid (4.02 g, 0.03 mol) in ethanol (100 mL) was prepared and added dropwise to the three-necked flask under constant stirring at 80°C. The mixture was refluxed for 3 h and a yellow Schiff base was obtained. Subsequently, DOPO (6.48 g, 0.03 mol) was added to the above mixture and stirred at 80°C for 3 h. Finally, the mixture was filtrated, the filter cake was washed three times with ethanol and dried at 100°C for 24 h to obtain a white powder (named as DT). The yield of DT was 71%. The synthesis route is shown in Scheme 1.

Preparation of BN@DT

BN@DT was prepared by ultra-sonication treatment and in-situ self-assembly method, and the complete synthesis process is depicted in Figure 1.

Typically, 4-aminobenzoic acid (2.74 g, 0.02 mol), BN (10.00 g) and ethanol (400 mL) were mixed in a 1000 mL three-necked flask and subjected to sonication at 40°C for 8 h. The ultra-sonication resulted in the separation of BN sheets and 4-aminobenzoic acid was immobilized onto the BN lamellae. After that, an ethanol solution of p-phthalaldehyde (1.34 g, 0.01 mol) was added dropwise to the mixture and the reaction was maintained at 80°C for 3 h. Afterwards, DOPO (4.32 g, 0.02 mol) was added and refluxed continuously at 80°C for 3 h. Finally, the mixture was filtrated, and the precipitates were washed three times with ethanol and dried at 100°C for 24 h to obtain a white powder (named as BN@DT).

Preparation of EP/BN@DT composites

In a typical procedure, BN@DT in various mass fractions (5%, 10%, and 15%) was uniformly dispersed into EP under strong magnetic stirring at 100°C. Subsequently, the curing agent DDM (one-quarter of the mass of EP) was fully dissolved in the mixture at 100°C. The resulted EP/BN@DT mixture was outgassed in vacuum at 110°C for 5 min, following which it was introduced into a Teflon mold. Finally, the composite was cured at the subsequent stepwise schedule of 120°C for 2 h and 150°C for 2 h. The aforementioned three composites were designated as EP/mBN@DT (m = 5, 10, 15). For comparison, EP/10BN/5DT, EP/15BN, and EP/15DT composites were prepared by the same process. The formulation of different samples is presented in Table 1.

Table 1.

The formulation of EP samples.

Sample code	BN@DT (wt%)	EP (wt%)	DT (wt%)	BN (wt%)
Pure EP	0	100	0	0
EP/5BN@DT	5	95	0	0
EP/10BN@DT	10	90	0	0
EP/15BN@DT	15	85	0	0
EP/10BN/5DT	0	85	5	10
EP/15BN	0	85	0	15
EP/15DT	0	85	15	0

Characterization

The surface morphology and microstructure of the samples were observed by using a field emission scanning electron microscopy (SEM, ZEISS Sigma 300) after a gold coating treatment (45 s, 10 mA, 3 kV), a field emission transmission electron microscopy (TEM, JEOL JEM-2100Plus) and an atomic force microscopy (AFM, Bruker Dimension Icon). Energy-dispersive X-ray spectroscopy (EDX) was conducted in parallel with the SEM measurements at an acceleration voltage of 15 kV.

Fourier transform infrared spectra (FTIR) were carried out by using the attenuated total reflection (ATR) method with an INVENIO FTIR spectrometer. X-ray diffraction (XRD) patterns were recorded by a D/max2550VB3+/PC X-ray diffractometer equipped with a voltage of 40 kV and Cu-Kα radiation. Thermogravimetric analysis (TGA, Netzsch TG 209 F3 Tarsus, Germany) was conducted under a nitrogen atmosphere at a heating rate of 20°C/min.

Thermogravimetric analyses/infrared spectrometry (TG-FTIR) was performed using a STA-2500 (Netzsch, German) thermogravimetric analyzer, which was interfaced to the IS-50 FTIR instruments (Thermo Fisher). The samples were scanned from 30 to 800°C at a heating rate of 10°C/min.

The limiting oxygen index (LOI) was determined by means of a HC-2 oxygen index instrument in accordance with the ASTM-D-2863 standard, with sample dimensions of 130 × 6.5 × 3.2 mm³. UL-94 vertical burning tests were conducted on a CFZ-3 apparatus (manufactured in Jiangning, China) in according to the ATSM D3801-20 standard. The sample size utilized in the test was 130 × 13 × 3 mm³. Cone calorimeter test (CCT) was performed on a FTT cone calorimeter (UK) at a heat flux of 35 kW/m², in accordance with the ISO 5660 standard. The sample size was 100 × 100 × 3 mm³.

The thermal conductivity was measured using hot disk method by a thermal conductivity meter (Hot Disk TPS2500S, Sweden). Raman spectra were obtained by LABIAN HR-EVOLUTION Raman Microscope (HORIBA Scientific, Japan) with a 532 nm laser. Tensile performance was ascertained by using a universal testing machine (TSE 105D, China) at a crosshead speed of 0.6 mm/min, in accordance with the ASTM D638-14 standard. The unnotched impact strength was measured using a pendulum impact tester (PIT501J, China) according to ASTM D256-10, with a specimen size of 80 × 10 × 4 mm³.

Results and discussion

Characterization of BN@DT

SEM, TEM and AFM images in Figure 2 revealed the morphology and microstructure of BN and BN@DT. A typical SEM image (Figure 2(a)) illustrates that the untreated BN presents stacked sheet structure with smooth surface and large thickness due to the close stacking of multiple layers. In contrast, the functionalized BN can be exfoliated to few-layer flakes via ultra-sonication and modification with the flame retardant DT, and the surface exhibits some characteristic rod-like attachments of DT (Figure 2(c)). In addition, the EDX mappings (Figure 2(d)–(h)) also confirm the presence of B, C, N, O, and P elements in BN@DT. Evidence from TEM images can further demonstrate the reduction in the thickness of the functionalized BN and the immobilization of DT on BN sheets (Figure 2(i) and (l)). Furthermore, AFM images reveal that untreated BN exhibits an average thickness of 200 nm and the lateral dimensions of approximately 2-6 μm (Figure 2(j)–(k)), while BN@DT shows an average thickness of 25 nm and a reduced lateral dimension (Figure 2(m) and (n)). The observed morphological and structural changes confirm the effective separation and dispersion of BN sheets, as well as the successful immobilization of DT on the surface of BN.

The FTIR spectra of BN, DT, and BN@DT are presented in Figure 3(a). For BN, two distinct peaks at approximately 1358 cm⁻¹ and 806 cm⁻¹ are attributed to the in-plane B-N stretching vibration and the out-of-plane B-N-B bending deformation, respectively.^32,33 For DT, the peaks at 3379 cm⁻¹ and 1664 cm⁻¹ correspond to the characteristic peak of hydroxyl (-OH) and the stretching vibrations of N-H³⁴; the peaks at 1602 cm⁻¹ (P-Ph), 1236 cm⁻¹ (P=O), and 1112 cm⁻¹ (P-O) are typical characteristic peaks of DOPO derivatives³⁵; the absorption peaks at 927 cm⁻¹ and 755 cm⁻¹ are ascribed to the P-O-Ph structure.³⁶ In addition, the FTIR spectrum of BN@DT exhibits the major characteristic peaks of both BN and DT, thereby confirming the successful combination of BN and DT.

XRD test was employed to elucidate the crystal structure of both BN and BN@DT. As shown in Figure 3(b), the characteristic diffraction peaks of BN are observed at 26.7, 41.5, 43.8, 50.1, and 55.1°, which correspond to the (002), (100), (101), (102), and (004) lattice planes, respectively.³⁰ After the functionalization, the basic crystal structure of BN remains largely unaltered, and BN@DT exhibits the primary diffraction peaks characteristic of BN and DT. It is noteworthy that a considerable reduction in intensity is observed for the (002) and (100) peaks of BN@DT, indicating the much less lamellar stacking after functionalization.⁹ The aforementioned results suggest the successful immobilization of DT on the surface of BN.

The thermal decomposition of BN, DT, and BN@DT was investigated by TG test, as shown in Figure 3(c) and (d). For BN, there is a slight mass loss even at 800°C, indicating its excellent thermal stability. In the case of DT, the temperature at 5% mass loss (T_5%) is observed at 296°C, with a temperature at the maximum mass loss rate (T_max) of 317°C. In comparison, the T_5% and T_max of BN@DT reach 310°C and 380°C, respectively. Moreover, the residual masses at 800°C of BN, DT, and BN@DT are 98.4%, 37.9%, and 77.3%, respectively. This indicates that the mass percentage of DT in BN@DT is approximately 35%, which is almost in accordance with the predesigned content.

Flame retardancy of EP composites

LOI and UL-94 analysis

The limiting oxygen index (LOI) and UL-94 vertical burning test were employed to evaluate the combustion behavior of EP composites, and the results are illustrated in Figure 4(a). Relative to the highly combustible pure EP (LOI = 23%, no rating), the addition of 15 wt% BN results in a notable increase in the LOI (29%), but EP/15BN has no rating at all in the UL-94 test. In comparison, EP/15DT exhibits a higher LOI (41%) and a UL-94 V-0 rating. The EP/5BN@DT, EP/10BN@DT and EP/15BN@DT composites exhibit a LOI value of 29%, 34%, and 40%, respectively. Among them, EP/15BN@DT can pass the V-0 rating, indicating its excellent flame retardancy. In parallel, the EP/10BN/5DT composite, prepared by simple blending of BN and DT, only exhibits a LOI of 37% and a V-1 rating, which is significantly inferior to that of EP/15BN@DT.

Figure 5 shows the combustion process of pure EP, EP/15BN@DT, and EP/15BN in the UL-94 test. Pure EP and EP/15BN burn fiercely and do not self-extinguish after being ignited, accompanied by a considerable amount of smoke (Figure 5(a) and (b)). In comparison, the EP/15BN@DT quickly extinguishes within 4 s after the first ignition and rapidly extinguishes within 3 s after the second ignition (Figure 5(c)). It should be noted that EP/15BN@DT releases a large amount of non-combustible gases during combustion, effectively blowing out the flame, which is a typical combustion phenomenon of phosphorus-containing flame retardant.^37,38 The aforementioned results suggest that the BN@DT has high flame-retardant efficiency in EP.

Cone calorimeter evaluation

The combustion behavior of EP composites was further investigated by cone calorimeter test (CCT), with the results presented in Figure 4(b)–(d) and Table 2. The peak heat release rate (PHRR) and total heat release (THR) values of pure EP are 1131 kW/m² and 113.5 MJ/m², respectively. In comparison, the PHRR values of EP/15BN, EP/10BN/5DT, EP/15BN@DT, and EP/15DT are 1059 kW/m², 790 kW/m², 721 kW/m², and 578 kW/m², a reduction of 6.4%, 30.2%, 36.3%, and 48.8%, respectively. Correspondingly, the THR values of the composites as described above exhibit a decline of 23.4%, 27.5%, 31.9%, and 44.1%, respectively. The incorporation of DT results in a substantial reduction in heat release, suggesting that DT effectively suppresses fire propagation presumably due to its participation in the charring reaction of the EP matrix. Furthermore, it is notable that the PHRR and THR values of EP/15BN@DT are lower than those of EP/15BN or EP/10BN/5DT, indicating that BN@DT is more effective in reducing the heat release of EP than BN or the direct blending of BN and DT. In EP/15BN@DT, DT produces inert gases on heating to dilute flammable gases and the decomposition products (phosphorus-containing acids) promote the formation of dense char, while the sheet-like BN increases the strength of the dense char, resulting in the excellent flame retardancy of the composite.^7,39

Table 2.

Cone calorimeter test results for pure EP and EP composites.

Sample code	TTI (s)	PHRR (kW/m²)	THR (MJ/m²)	TSP (m²)	FPI (s·m²/kW)
Pure EP	120	1131	113.5	21.6	0.106
EP/15BN	96	1059	86.9	15.8	0.091
EP/10BN/5DT	98	790	82.3	15.4	0.124
EP/15BN@DT	97	721	77.3	13.2	0.135
EP/15DT	80	578	63.6	12.5	0.139

The total smoke production (TSP) serves as a crucial indicator for evaluating the smoke emission of polymer materials. EP/15BN, EP/10BN/5DT, EP/15BN@DT and EP/15DT display the TSP of 15.8 m², 15.4 m², 13.2 m² and 12.5 m², respectively, a reduction of 26.9%, 28.7%, 38.9%, and 42.1%, respectively compared with pure EP (21.6 m²). BN functioning as a physical barrier can effectively inhibit the release of smoke during the combustion of EP.⁴⁰ The significant reduction in TSP of EP/15BN@DT may be due to the formation of the good char layer.

The fire performance index (FPI), defined as the ratio of the time to ignition (TTI) to PHRR, is used to evaluate the potential fire hazard of materials.⁴¹ It is generally accepted that a higher FPI indicates a lower fire risk and better flame retardancy. As shown in Table 2, the FPI of pure EP is 0.106 s·m²/kW. The EP/15BN@DT and EP/15DT exhibit notably elevated FPI values of 0.135 and 0.139 s·m²/kW, respectively, confirming their better flame-retardant performance.

Table 3.

Thermal decomposition data of pure EP and its composites.

Sample code	T_5% (°C)	T_max (°C)	Char residue at 800°C (wt%)
Pure EP	383	395	14.1
EP/15DT	338	370	20.8
EP/15BN	376	394	27.4
EP/15BN@DT	355	384	24.9

Analysis of char residues

Figure 6 shows the digital photos and SEM images of pure EP and its composites after CCT. The pure EP is almost burnt out, with very few incomplete char residues (Figure 6a1–a2), indicating its poor flame retardancy. In contrast, the amount of char residue and the density of char layer increase significantly after the combustion of EP/15DT, with a char thickness of 4.3 cm (Figure 6b1–b2). This phenomenon suggests that DT can facilitate the formation of dense char layer, which subsequently inhibits the exchange of oxygen and heat between the combustion zone and the condensed phase, thereby slowing down the decomposition of EP composites.⁴² As shown in Figure 6c1–c2, EP/15BN has a disrupted and discontinuous char layer, but the thickness and density of the char are both improved compared to pure EP. In the case of EP/10BN/5DT, the char layer is more complete than that of EP/15BN (Figure 6d1–d2). It is notable that the EP/15BN@DT exhibits a more continuous and dense char, with a thickness of 3.2 cm (Figure 6e1–e2).

In the SEM image, the char residue of pure EP exhibits many large pores, and thus cannot provide effective insulation against the heat and flammable gases (Figure 6a3). In contrast, the char layers of EP/15DT, EP/15BN and EP/10BN/5DT are more complete and continuous, with only few micro-pores and cracks (Figure 6b3–d3). In the case of EP/15BN@DT, the char layer is much more intact and dense, with almost no visible defects (Figure 6e3). Reasonably, this kind of char structure serves to impede the release of gaseous fuels into the gaseous phase, while simultaneously reducing the efficiency of heat and oxygen exchange between the condensed and vapor phases, effectively preventing further decomposition of the underlying material, as evidenced by the great reductions in PHRR, THR and TSP values of EP/15BN@DT.⁴³

Raman spectroscopy is usually employed to reveal the graphitization degree of char residue.⁴⁴ Figure 7 shows Raman spectra of the char residue of pure EP and its composites. For pure EP, there are two obvious peaks at approximately 1351 cm⁻¹ and 1589 cm⁻¹ (Figure 7(a)), corresponding to the disordered (D band) and ordered (G band) structures. The graphitization degree of char residue is usually evaluated by the ratio of the integrated intensity of the D and G bands (I_D/I_G).⁴⁵ In general, a smaller I_D/I_G value implies a higher degree of graphitization and a more stable char layer, which can act as a barrier to prevent further combustion of unburned polymer matrix.⁴³ The I_D/I_G value of pure EP is 3.5, and the values of EP/15BN and EP/15BN@DT decrease to 3.1 and 2.2, respectively, indicating that EP/15BN@DT after combustion can form a more stable char layer than EP/15BN or pure EP, and thus achieve better flame-retardant performance.

Thermal degradation of EP composites

The thermal degradation of EP composites was evaluated by TG under nitrogen atmosphere, and the results are presented in Figure 8 and Table 3. For all samples, there is only one peak (Figure 8(b)), which implies the existence of a single degradation process between 300°C and 500°C. It is clear that EP/15DT exhibits a 45°C lower T_5% and a 25°C lower T_max in comparison to EP, which is ascribed to the decomposition of DT. In contrast, the T_5% and T_max of EP/15BN are essentially identical to those of pure EP, indicating that the influence of BN on EP degradation is insignificant. The T_5% and T_max of EP/15BN@DT are 28°C and 11°C lower than those of pure EP, respectively. The quantity of char residue at 800°C increases from 14.1% of EP to 20.8% of EP/15DT, indicating the well char-forming ability of DT.⁴⁶ For EP/15BN, the residual mass at 800°C is 27.4%, which is attributed to the excellent thermal stability of BN. The char residue of EP/15BN@DT is 24.9%, indicating the formation of protective chars and a reduction in heat and fuel generated from matrix degradation during combustion, thereby enhancing the flame retardancy of EP composites.

In order to gain deeper insight into the flame-retardant mechanism, TG-FTIR was utilized to investigate the pyrolysis products of pure EP and EP/15BN@DT, and the results are displayed in Figure 9. The 3D FTIR spectrum of EP/15BN@DT exhibits a significant decrease in the absorbance intensities relative to pure EP, suggesting a dramatically reduced presence of volatile products (Figure 9(a) and (b)). As presented in Figure 9(c), the major volatile pyrolysis products of pure EP and EP/15BN@DT detected at T_max include H₂O (3900–3500 cm⁻¹), hydrocarbons (3100–2800 cm⁻¹), CO₂ (2360, 2337 cm⁻¹), carbonyl compounds (1705 cm⁻¹), aromatic compounds (1600-1500 cm⁻¹, 829 cm⁻¹) and ester compounds (1174 cm⁻¹).^7,47 The presence of BN@DT significantly decreases the amounts of gaseous products of the EP composite. These volatile organic compounds not only release large amounts of heat during combustion, but also form some toxic fumes. Consequently, the reduction in the volatile pyrolysis compounds indicates a reduction in “fuel,” ultimately leading to a reduction in heat release and smoke production, as evidenced by the great reductions in PHRR, THR and TSP values during the CCT.

From the above comprehensive analysis, a possible mechanism for the enhanced fire safety of EP/BN@DT can be explained by the barrier effect of BN and the catalytic carbonization effect of DT. During combustion, DT decomposes to release some phosphorus-containing acids that catalyze the carbonization of EP matrix, thereby promoting the formation of a protective char layer.³⁶ Simultaneously, the presence of non-combustible BN sheets contributes to the formation of a more rigid and compact char layer, which plays some flame-retardant role in the EP/BN@DT composites.⁴⁸ The physical barrier effect of BN and the catalysis carbonization of DT make the char layer more complete and stronger, thus effectively preventing the transfer of heat, oxygen and fuel between the gaseous and condensed phases.⁴⁹

Thermally conductive performance of EP composites

Figure 10(a) illustrates the in-plane thermal conductivity (λ) of pure EP and EP composites at 25°C. The pure EP has a relatively low λ (0.21 Wm⁻¹K⁻¹) due to the large thermal resistance generated by the random orientation of amorphous molecular chains.^5,50 The λ of EP/15DT is slightly improved, with a value of approximately 0.27 Wm⁻¹K⁻¹. The incorporation of an appropriate quantity of BN sheets into the EP matrix facilitates the formation of inorganic filler networks, which can form the interconnected heat conduction paths, resulting in the discernible enhancement in the λ of EP/BN composites.⁹ EP/10BN/5DT, EP/15BN@DT and EP/15BN display the λ values of 0.44 Wm⁻¹K⁻¹, 0.52 Wm⁻¹K⁻¹, and 0.61 Wm⁻¹K⁻¹, respectively, representing an improvement of 110%, 148%, and 190%, respectively, compared with that of pure EP. Notably, EP/15BN@DT exhibits a higher λ than EP/10BN/5DT, which is ascribed to the enhanced dispersion of BN@DT in EP and the augmented compatibility with EP. On the one hand, the well-dispersed BN@DT facilitates the formation of additional heat conduction pathways within the matrix, thereby enhancing the efficiency of phonon transmission. On the other hand, the good compatibility greatly reduces the interfacial thermal resistance between the BN@DT and EP.³⁰ It is reasonably deduced that the flame-retardant functionalization of BN represents a viable method for enhancing the thermal conductivity of EP/BN composites.

To further investigate the thermal management capability of EP composites, pure EP and EP/15BN@DT were subjected to a 150°C heat treatment in an oven, followed by a natural cooling process. Subsequently, the temperature changes on the surface of the samples were measured, and the results are shown in Figure 10(b). Following the commencement of the cooling process, the surface temperatures of the EP/15BN@DT are all lower than those of pure EP at a fixed time during the cooling process, with a mean difference of approximately 10°C. In particular, the surface temperature of EP/15BN@DT at 40 s is 77°C, a 14°C lower than that of pure EP (91°C). The above results demonstrate that EP/BN@DT composites can dissipate the excessive heat generated by electronic devices effectively.

Mechanical performance of EP composites

Figure 11 illustrates the tensile and impact strengths of EP composites. Pure EP has a tensile strength of 51.8 MPa and an impact strength of 9.4 KJ/m². As the content of BN@DT increases, EP/5BN@DT, EP/10BN@DT and EP/15BN@DT exhibit a slight decrease trend in tensile strength, with values of 48.6 MPa, 46.5 MPa, and 44.5 MPa, respectively. Meanwhile, the impact strengths of the composites increase slightly and then decrease a bit with an increase in the BN@DT content. However, both the tensile and impact strengths of EP/15BN@DT are higher than the strengths of EP/15BN or EP/10BN/5DT at a fixed loading (15 wt%). The reason for the higher mechanical strengths of EP/15BN@DT is possibly because the immobilization of DT on the surface of BN can facilitate the dispersion of BN@DT in the EP matrix and enhance their interfacial interaction.

Morphology of EP composites

To further investigate the dispersion and compatibility of BN or modified BN in EP matrix, the fracture morphology and structure of EP composites were observed by SEM. As shown in Figure 12(a), the cross section of Pure EP exhibits a relatively smooth and regular appearance, a typical characteristic of brittle failure. For EP/15BN (Figure 12(b)), there is apparent aggregation and phase separation of BN, indicating the poor dispersion and weak interfacial interaction of BN within the EP matrix, which substantially deteriorates the mechanical properties of EP. Figure 12(d) shows that the aggregation of BN is still evident after a simple blending of BN and DT in the EP matrix. In contrast, DT is well dispersed in EP/15DT composites without obvious agglomeration (Figure 12(c)). The good dispersion of DT is probably attributed to the strong interactions between its abundant reactive groups (such as -NH, -COOH) and the EP matrix.^28,43 Due to the immobilization of DT on the surface of BN, the interfacial interaction between BN@DT and the EP matrix is enhanced so that there is no apparent filler agglomeration throughout the cross section of EP/BN@DT composites (Figure 12(e)–(g)). In addition, EDX images confirm the uniform dispersion of BN@DT in the fracture of EP/15BN@DT (Figure 12(h)). These results suggest that the functionalization of the BN surface by DT improves the interfacial interactions between BN and the EP matrix, which may lead to the improvement of the flame retardancy, thermally conductive performance and mechanical properties of EP composites.

Conclusions

In this work, a phosphorus-containing flame retardant (DT) functionalized BN (BN@DT) was prepared based on ultra-sonication treatment and in-situ self-assembly method, wherein the DT was immobilized on the surface of BN. As expected, the incorporation of DT can improve the dispersion of BN in EP and the interfacial interaction with EP matrix. It is found that the EP composite containing 15% BN@DT (EP/15BN@DT) can pass the UL-94 V-0 rating with a LOI value of 40%. The PHRR, THR and TSP values of EP/15BN@DT decrease by 36.3%, 31.9% and 38.9%, respectively compared to the values of pure EP. Simultaneously, the thermal conductivity of EP/15BN@DT is 148% higher than that of pure EP. Furthermore, the tensile and impact strengths of EP/15BN@DT decrease slightly in comparison to pure EP. In short, this work offers a promising method for the preparation of the thermally conductive EP composites with an excellent flame retardancy.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was financially supported by the National Natural Science Foundation of China (No. 21975185).

ORCID iD

Zhengzhou Wang

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.*

References

Luo

Zhao

Bai

, et al. Peak-type charge barriers enabling high surface-insulating performance of epoxy composites. Compos Sci Technol 2022; 229: 109706. DOI: 10.1016/j.compscitech.2022.109706.

Leung

. Thermally conductive polymer composites and nanocomposites: processing-structure-property relationships. Compos Part B-Eng 2018; 150: 78–92. DOI: 10.1016/j.compositesb.2018.05.056.

Yang

Chen

Zhao

, et al. Enhancing dielectric strength of thermally conductive epoxy composites by preventing interfacial charge accumulation using micron-sized diamond. Compos Sci Technol 2022; 221: 109178. DOI: 10.1016/j.compscitech.2021.109178.

, et al. Facile construction of a flexible film with ultrahigh thermal conductivity and excellent flame retardancy for a smart fire alarm. Chem Mater 2021; 33: 3228–3240. DOI: 10.1021/acs.chemmater.1c00113.

Feng

Han

Wang

, et al. Multiple synergistic effects of graphene-based hybrid and hexagonal born nitride in enhancing thermal conductivity and flame retardancy of epoxy. Chem Eng J 2020; 379: 122402. DOI: 10.1016/j.cej.2019.122402.

Zhang

Liu

Zhang

, et al. High-temperature resistance and excellent electrical insulation in epoxy resin blends. J Appl Polym Sci 2023; 140: 54359. DOI: 10.1002/app.54359.

Zhang

Sui

, et al. Synergistic effect of novel ionic liquid/graphene complex on the flame retardancy of epoxy nanocomposites. Carbon Lett 2023; 33: 501–516. DOI: 10.1007/s42823-022-00440-9.

Awais

Chen

Hong

, et al. Synergistic effects of micro-hBN and core-shell nano-TiO₂@SiO₂ on thermal and electrical properties of epoxy at high frequencies and temperatures. Compos Sci Technol 2022; 227: 109576. DOI: 10.1016/j.compscitech.2022.109576.

Han

Zhao

Feng

, et al. Highly flame-retardant epoxy-based thermal conductive composites with functionalized boron nitride nanosheets exfoliated by one-step ball milling. Chem Eng J 2021; 407: 127099. DOI: 10.1016/j.cej.2020.127099.

10.

Chen

Xue

, et al. Construction of 3D boron nitride nanosheets/silver networks in epoxy-based composites with high thermal conductivity via in-situ sintering of silver nanoparticles. Chem Eng J 2019; 369: 1150–1160. DOI: 10.1016/j.cej.2019.03.150.

11.

Liu

, et al. Casein phosphopeptide-biofunctionalized graphene oxide nanoplatelets based cellulose green nanocomposites with simultaneous high thermal conductivity and excellent flame retardancy. Chem Eng J 2020; 382: 122733. DOI: 10.1016/j.cej.2019.122733.

12.

Feng

Xue

, et al. Simultaneous improvement in the flame resistance and thermal conductivity of epoxy/Al₂O₃ composites by incorporating polymeric flame retardant-functionalized graphene. J Mater Chem A 2017; 5: 13544–13556. DOI: 10.1039/c7ta02934a.

13.

Zheng

Liu

, et al. Thermally conductive composites based on hexagonal boron nitride nanosheets for thermal management: fundamentals to applications. Compos Part A-Appl S 2023; 169: 107533. DOI: 10.1016/j.compositesa.2023.107533.

14.

Chen

Huang

Sun

, et al. Highly thermally conductive yet electrically insulating polymer/boron nitride nanosheets nanocomposite films for improved thermal management capability. ACS Nano 2019; 13: 337–345. DOI: 10.1021/acsnano.8b06290.

15.

Guerra

Wan

McNally

. Thermal conductivity of 2D nano-structured boron nitride (BN) and its composites with polymers. Prog Mater Sci 2019; 100: 170–186. DOI: 10.1016/j.pmatsci.2018.10.002.

16.

Wang

Sun

. Graphene, hexagonal boron nitride, and their heterostructures: properties and applications. RSC Adv 2017; 7: 16801–16822. DOI: 10.1039/c7ra00260b.

17.

Zhou

Zhang

, et al. 2D boron nitride nanosheets for smart thermal management and advanced dielectrics. Adv Mater Interfac 2022; 9: 2200610. DOI: 10.1002/admi.202200610.

18.

Song

Liu

, et al. Two-dimensional materials for thermal management applications. Joule 2018; 2: 442–463. DOI: 10.1016/j.joule.2018.01.006.

19.

Jiang

BinteTouhid

Sun

, et al. Preparation of functionalized boron nitride sheets/epoxy resin composites by using a green and efficient approach for elevated thermal conductivity. Polym Compos 2023; 44: 6826–6838. DOI: 10.1002/pc.27600.

20.

Hong

Jung

Lee

, et al. Anisotropic thermal conductive composite by the guided assembly of boron nitride nanosheets for flexible and stretchable electronics. Adv Funct Mater 2019; 29: 1902575. DOI: 10.1002/adfm.201902575.

21.

Xiao

Guo

Tang

, et al. Epoxy composite with significantly improved thermal conductivity by constructing a vertically aligned three-dimensional network of silicon carbide nanowires/boron nitride nanosheets. Compos Part B-Eng 2020; 187: 107855. DOI: 10.1016/j.compositesb.2020.107855.

22.

Han

Shi

Wang

, et al. Nest-like hetero-structured BNNS@SiCnws fillers and significant improvement on thermal conductivities of epoxy composites. Compos Part B-Eng 2021; 210: 108666. DOI: 10.1016/j.compositesb.2021.108666.

23.

Chen

Liu

, et al. Simultaneous production and functionalization of boron nitride nanosheets by sugar-assisted mechanochemical exfoliation. Adv Mater 2019; 31: 1804810. DOI: 10.1002/adma.201804810.

24.

Lei

, et al. Surface modifications of boron nitride nanosheets for poly(vinylidene fluoride) based film capacitors: advantages of edge-hydroxylation. J Mater Chem A 2019; 7: 7664–7674. DOI: 10.1039/c9ta00616h.

25.

Ding

Zhao

. High-yield synthesis of extremely high concentrated and few-layered boron nitride nanosheet dispersions. 2D Mater 2018; 5: 045015. DOI: 10.1088/2053-1583/aad51a.

26.

Jiang

Shi

Feng

, et al. Enhanced thermal conductivity and ideal dielectric properties of epoxy composites containing polymer modified hexagonal boron nitride. Compos Part A-Appl S 2018; 107: 657–664. DOI: 10.1016/j.compositesa.2018.02.016.

27.

Liang

Zhao

, et al. Highly thermally conductive flame-retardant epoxy nanocomposites with reduced ignitability and excellent electrical conductivities. Compos Sci Technol 2017; 139: 83–89. DOI: 10.1016/j.compscitech.2016.12.015.

28.

Chen

Wang

. A copper organic phosphonate functionalizing boron nitride nanosheet for PVA film with excellent flame retardancy and improved thermal conductive property. Compos Part A-Appl S 2022; 153: 106738. DOI: 10.1016/j.compositesa.2021.106738.

29.

Yao

Qian

Qiu

, et al. Flame retardant and toughening behaviors of bio-based DOPO-containing curing agent in epoxy thermoset. Polym Adv Technol 2020; 31: 461–471. DOI: 10.1002/pat.4782.

30.

Yang

Guo

Zhang

, et al. Significant improvement in the flame retardancy and thermal conductivity of the epoxy resin via constructing a branched flame retardant based on SI-ATRP initiated by dopamine-modified boron nitride. Ind Eng Chem Res 2022; 61: 8031–8042. DOI: 10.1021/acs.iecr.2c01044.

31.

Hong

Xing

Guo

, et al. Fabrication of thermally conductive boron nitride nanosheets-conjugated poly (3-hexylthiophene) composite with enhanced photothermal property. Mater Lett 2023; 333: 133673. DOI: 10.1016/j.matlet.2022.133673.

32.

Yan

Dai

Gao

, et al. Ultrahigh-aspect-ratio boron nitride nanosheets leading to superhigh in-plane thermal conductivity of foldable heat spreader. ACS Nano 2021; 15: 6489–6498. DOI: 10.1021/acsnano.0c09229.

33.

Lee

Park

, et al. Scalable exfoliation process for highly soluble boron nitride nanoplatelets by hydroxide-assisted ball milling. Nano Lett 2015; 15: 1238–1244. DOI: 10.1021/nl504397h.

34.

Osipov

Aleksenskiy

Shames

, et al. Infrared absorption study of surface functional groups providing chemical modification of nanodiamonds by divalent copper ion complexes. Diam Relat Mater 2011; 20: 1234–1238. DOI: 10.1016/j.diamond.2011.07.008.

35.

Zhang

, et al. Strong yet tough epoxy with superior fire suppression enabled by bio-based phosphaphenanthrene towards in-situ formed Diels-Alder network. Compos Part B-Eng 2023; 251: 110490. DOI: 10.1016/j.compositesb.2022.110490.

36.

Zhang

Yang

Wang

, et al. A DOPO based reactive flame retardant constructed by multiple heteroaromatic groups and its application on epoxy resin: curing behavior, thermal degradation and flame retardancy. Polym Degrad Stabil 2019; 167: 10–20. DOI: 10.1016/j.polymdegradstab.2019.06.020.

37.

Wang

Xing

, et al. Two-dimensional cardanol-derived zirconium phosphate hybrid as flame retardant and smoke suppressant for epoxy resin. Polym Degrad Stabil 2018; 151: 172–180. DOI: 10.1016/j.polymdegradstab.2018.03.006.

38.

Han

Hao

Chen

, et al. Schiff base derivative flame-retardant endowing epoxy resin with toughness, transparency, and high flame retardancy. Eur Polym J 2024; 214: 113114. DOI: 10.1016/j.eurpolymj.2024.113114.

39.

Shi

Wei

Gao

. Enhanced flame retardancy of epoxy composites containing melamine polyphosphate-modified boron nitride. J Appl Polym Sci 2022; 139: 53215. DOI: 10.1002/app.53215.

40.

Jin

Yuan

Liang

, et al. Multifunctional cyclotriphosphazene/hexagonal boron nitride hybrids and their flame retarding bismaleimide resins with high thermal conductivity and thermal stability. ACS Appl Mater Interfaces 2014; 6: 14931–14944. DOI: 10.1021/am502364k.

41.

Zhou

Qiu

, et al. Synthesis of star-shaped allyl phosphazene small molecules for enhancing fire safety and toughness of high performance BMI resin. Chem Eng J 2021; 425: 130655. DOI: 10.1016/j.cej.2021.130655.

42.

Guo

Zhao

Wang

, et al. Multifunctional epoxy composites with highly flame retardant and effective electromagnetic interference shielding performances. Compos Part B-Eng 2020; 192: 107990. DOI: 10.1016/j.compositesb.2020.107990.

43.

Chen

Luo

, et al. A universal strategy toward flame retardant epoxy resin with ultra-tough and transparent properties. Polym Degrad Stabil 2022; 205: 110132. DOI: 10.1016/j.polymdegradstab.2022.110132.

44.

Yang

Huo

Wang

, et al. A highly fire-safe and smoke-suppressive single-component epoxy resin with switchable curing temperature and rapid curing rate. Compos Part B-Eng 2021; 207: 108601. DOI: 10.1016/j.compositesb.2020.108601.

45.

Zheng

Zhou

. A novel exploration of metal-organic frameworks in flame-retardant epoxy composites. J Therm Anal Calorim 2019; 138: 905–914. DOI: 10.1007/s10973-019-08267-9.

46.

Wang

Tan

, et al. A novel and feasible approach for one-pack flame-retardant epoxy resin with long pot life and fast curing. Chem Eng J 2018; 337: 30–39. DOI: 10.1016/j.cej.2017.12.086.

47.

Chen

Liu

Zhang

, et al. An iron phenylphosphinate@graphene oxide nanohybrid enabled flame-retardant, mechanically reinforced, and thermally conductive epoxy nanocomposites. Chem Eng J 2023; 454: 140424. DOI: 10.1016/j.cej.2022.140424.

48.

Wang

Yin

, et al. Hexagonal boron nitride@ZnFe₂O₄ hybrid nanosheet: an ecofriendly flame retardant for polyvinyl alcohol. J Solid State Chem 2020; 287: 121366. DOI: 10.1016/j.jssc.2020.121366.

49.

Wang

. In-situ preparation of layered zinc N, N′-piperazine (bismethylene phosphonate) functionalizing reduced graphene oxide for epoxy resin with reduced fire hazards and improved thermal/mechanical properties. Compos Part A-Appl S 2021; 149: 106588. DOI: 10.1016/j.compositesa.2021.106588.

50.

Burger

Laachachi

Ferriol

, et al. Review of thermal conductivity in composites: mechanisms, parameters and theory. Prog Polym Sci 2016; 61: 1–28. DOI: 10.1016/j.progpolymsci.2016.05.001.