Covalent assembly of MCM-41 nanospheres on graphene oxide for improving fire retardancy and mechanical property of epoxy resin

Covalent assembly of MCM-41 nanospheres on graphene oxide for improving fire retardancy and mechanical property of epoxy resin

Composites Part B 138 (2018) 101–112 Contents lists available at ScienceDirect Composites Part B journal homepage: www.elsevier.com/locate/composite...

2MB Sizes 0 Downloads 0 Views

Composites Part B 138 (2018) 101–112

Contents lists available at ScienceDirect

Composites Part B journal homepage: www.elsevier.com/locate/compositesb

Covalent assembly of MCM-41 nanospheres on graphene oxide for improving fire retardancy and mechanical property of epoxy resin

T

Zhi Lia, Alejandro Jiménez Gonzáleza,b, Vignesh Babu Heeralala, De-Yi Wanga,∗ a b

IMDEA Materials Institute, C/Eric Kandel, 2, 28906 Getafe, Madrid, Spain Escuela Superior de Ciencias Experimentales y Tecnología (ESCET), Universidad Rey Juan Carlos, C/Tulipán, s/n, 28933 Móstoles, Madrid, Spain

A R T I C L E I N F O

A B S T R A C T

Keywords: Polymer-matrix composites (PMCs) Fire retardancy Assembly Thermosetting resin

In the study, the hierarchical nanohybrid ([email protected]) with mesoporous MCM-41 nanospheres covalent assembling on graphene oxide (GO) nanosheets was successfully prepared, aiming to improve fire retardancy of epoxy resin (EP). Fourier transformation infrared spectra (FTIR), Raman spectra, X-ray diffraction (XRD), X-ray photoelectron spectra (XPS), scanning electron microscopy (SEM) and transmission electron microscopy (TEM) verified [email protected] The nano-dispersion of [email protected] in EP was certified via TEM and XRD. Fire-retardant investigation (cone calorimeter test) revealed that EP nanocomposite with 2 wt% [email protected] possessed a 40.0% drop of peak heat release rate (pHRR) and suppressed smoke production relative to EP/GO. Intriguingly, EP/[email protected] demonstrated lower pHRR than EP/GO-MCM-41 (direct addition of GO and MCM-41). TG analysis indicated [email protected] imparted EP with significantly reduced degradation rate and increased char yield, accompanied by reduced volatile in thermogravimetric analysis couple with Fourier transformation infrared spectra (TG-FTIR) and improved aromatization in variable-temperature FTIR. The impact strength of EP/[email protected] was enhanced by 38.8% compared with EP. Contrastive analysis proposed the mechanism of neighboring synergy of GO and MCM-41, which involved in GO nanosheets wrapped by enormous carbonaceous stuffs due to MCM-41 neighboring catalysis. In perspective, hierarchical nano-assembly with neighboring synergy offered a viable approach for fire-safe polymers.

1. Introduction Epoxy resin (EP) nanocomposite has obtained intensive academic and industrialized interests recently owing to improved mechanical performance, fire retardancy and dimensional stability in comparison to neat EP [1–3]. Particularly in certain high-demanding circumstances, fire retardancy of EP is an indispensable requirement, which continuously drove the development of optimised nanofillers to generate fire-retardant EP nanocomposites. Generally, the factors influencing fire retardancy will be fully considered, which included particle type, particle size, particle shape, inter-particle synergistic effect and inter-particle spatial relationship in matrix [4–6]. EP nanocomposite possessed various fire-retardant mechanisms of barrier effect [7–11], catalyticcharring effect [12] and moisture-releasing effect [13–15]. In the sense, it was reasonable that the synergistic combination of different nanoparticles with preferential fire-retardant mechanism will promote fire retardancy more pronouncedly in contrast to single-component EP nanocomposites. Intriguingly, the combination of different nanofillers in spatial relationship of encapsulation or decoration indeed yielded



Corresponding author. E-mail address: [email protected] (D.-Y. Wang).

https://doi.org/10.1016/j.compositesb.2017.11.001 Received 7 May 2017; Received in revised form 29 August 2017; Accepted 2 November 2017 Available online 11 November 2017 1359-8368/ © 2017 Elsevier Ltd. All rights reserved.

excellent fire retardancy [16]. The reason was probably attributed to mutually promoted dispersion of nanofillers in polymer matrix or/and other unclear synergistic fire retardancy process. Basically, the spatial relationship of different nanoparticles with preferential fire-retardant mechanism in EP matrix is probably critical to determine ultimate fire retardancy. Currently, the study about how to design an efficient spatial relationship of appropriate nanofillers with preferential fire-retardant mechanism was still lacking and ongoing. Graphene oxide (GO) is a typical two-dimensional carbon nanomaterial with carbon atoms distributed in lamellar form. Compared with graphene nano-lamellas, abundant oxygen-containing groups are present in GO nanosheets, which provide enormous sites for further functionalization [17]. Actually, GO inhibited combustion behavior mainly through the barrier effect [18]. However, due to the ease of thermal decomposition of GO nanosheets, its nanocomposites merely demonstrated slightly enhanced or even deteriorated fire retardancy [19]. Therefore, further modification of GO nanosheets was necessary to weaken or avoid thermal decomposition during combustion. One feasible approach was to wrap protective layers over GO nanosheets

Composites Part B 138 (2018) 101–112

Z. Li et al.

3 mL of 30 wt% H2O2 were then poured into the system, giving rise to enormous bubbles. After the cooling, the mixture was purified using centrifugation, filtration and washing until pH value reached around 7. Finally, GO nanosheets were collected as the black solid after vacuum drying overnight.

[20]. The mesoporous MCM-41 with pore size between 2 nm and 50 nm draws extensive focuses owing to the ease of preparation and wide utilizations in catalysis fields [21]. It was reported that MCM-41 possessed fire-retardant reinforcement in intumescent systems based on the catalytic-charring effect [22,23]. Specially, nanoscale MCM-41 demonstrated more excellent catalytic-charring behavior than respective microscale counterpart due to more catalytic sites. Accordingly, it was promising to combine preferential fire-retardant mechanism of GO nanosheets and MCM-41 nanospheres through the assembly of MCM-41 on GO, accompanied by mitigating the drawback of MCM-41 to deteriorate char structure and GO to thermally degrade during combustion. In the study, mesoporous MCM-41 nanospheres were covalently assembled on GO sheets to fabricate hierarchical nanohybrid ([email protected] MCM-41), aiming to impart EP with enhanced fire retardancy. Fourier transformation infrared spectra (FTIR), X-ray diffraction (XRD), X-ray photoelectron spectra (XPS), scanning electron microscopy (SEM), transmission electron microscopy (TEM) and Raman spectrum were employed to characterize targeted products. Afterwards, the obtained [email protected] was incorporated into EP. The nano-dispersion structure and spatial relationship of GO and MCM-41 were investigated by XRD and TEM. Meanwhile, thermogravimetric (TG) analysis, variable-temperature FTIR (vt FTIR) and TG-FTIR were used to analyze the thermal degradation behavior and volatiles evolutions. Fire retardancy was investigated in cone calorimeter test (CCT), accompanied by char analysis with SEM and Raman spectra. Finally, dynamic mechanical analysis and non-notched impact test were performed.

2.4. Covalent assembly of [email protected] 0.35 g GO and 2 g NH2-MCM-41 were charged into 400 mL ethanol, followed by heating to 80 °C and refluxing for 24 h. During the reaction, the surfactant CTAB in the mesopores was dissolved in ethanol very easily. Finally, the dry product was collected after filtration and washing for several times prior to vacuum drying overnight. The schematic illustration was demonstrated in Fig. 1 (b). It was notable that the mass ratio of GO and NH2-MCM-41 was selected based on the approximate calculation (mass ratio of GO and NH2-MCM-41, around 1:20) according to three-layered sandwich model and the previous literature [26]. 2.5. Preparation of EP nanocomposites EP nanocomposites with NH2-MCM-41, GO, [email protected] and direct mixture of GO and amino grafting MCM-41 (GO-MCM-41) were fabricated at weight fraction of 2 wt% separately. In detail, the nanofiller was incorporated into epoxy monomer at 50 °C under vigorous stirring, followed by processing using three-roll milling machine to attain the good dispersion. Afterwards, the curing agent DDS was added to the resultant mixture at constant 125 °C with strong stirring to guarantee complete dissolution prior to transferring into the vacuum oven to degas at 115 °C for 5min. Subsequently, the mixture was casted into the preheated silicon mold to fabricate different specimen based on the procedure: 160 °C for 1 h, 180 °C for 2 h and 200 °C for 1 h. The neat EP was prepared in the same procedure. The acquired products were labelled as the EP, EP/GO, EP/MCM-41, EP/[email protected] and EP/GOMCM-41 correspondingly.

2. Experimental 2.1. Raw materials Graphite, condensed sulphuric acid (H2SO4), hydrogen peroxide (H2O2, 30 wt%), tetraethylorthosilicate (TEOS), potassium permanganate (KMnO4), sodium nitrate (NaNO3), hexadecyltrimethylammonium bromide (CTAB), sodium hydroxide (NaOH) and γ-aminopropyltriethoxysilane (KH-550) were purchased from Sigma-Aldrich. Ethanol (96% w/w) was acquired from Panreac Company. Epoxydharz C was obtained from Faserverbundwerkstoffe Composite Technology Company. 4,4′-diaminodiphenylmethane (DDS) and the control sample (graphene) was provided by TCI EUROPA. The deionized water was produced by our institute.

2.6. Instrumental FTIR spectra were performed using FTIR spectrometer (Nicolet iS50). Variable-temperature FTIR (vt FTIR) spectra were carried out on the identical equipment with temperature heating program (increasing from 50 °C to 400 °C at 10 °C/min and then keeping at 400 °C for 10min). Raman spectra were operated on RAMAN micro-spectroscopy system (Renishaw PLC). Wire 4.1 software was employed to fit and calculate the peak area size. XRD spectra were carried out on XPERTPRO diffractometer with Ni filter and Cu Kα radiation (λ = 0.154). XPS spectra were recorded on VG ESCALAB MK II spectrometer with Al K α X-ray radiation. TEM observation of EP nanocomposites was conducted on FEG S/TEM microscopy (Talos F200X, FEI) equipped with dispersive X-ray spectroscopy (EDS) at the accelerating voltage of 80 kV. The ultrathin sections were prepared by ultramicrotomy. Besides, the nanohybrids were analyzed in TEM observations (TEM, EM912, Zeiss) with acceleration voltage of 220 kV N2 sorption measurements were carried out on Autosorb-1 (Quantachrome, USA). BET (Brunauer-EmmettTeller) calculation of nitrogen isotherms at −196 °C was used to confirm specific surface size of MCM-41. Pore size and distribution were determined using BJH (Barrett-Joyner-Halenda) formula. Pore volume was obtained at the relative pressure (P/P0) of 0.98. SEM characterization was conducted on the equipment (SEM, EVO MA15, Zeiss). In order to probe MCM-41 nanospheres, FIB-FEGSEM dual-beam microscope (Helios NanoLab 600i, FEI) was employed. Thermal and thermooxidation degradation were investigated using thermogravimetric analyzer (Q50, TA Instruments) with heating rate of 10 °C/min from ambient temperature to 700 °C. TG-FTIR characterization was conducted using FTIR spectrometer (Nicolet iS50) connected with thermogravimetric analyzer (Q50, TA Instruments) through heated

2.2. Preparation of mesoporous amino grafting MCM-41 nanospheres Mesoporous amino grafting MCM-41 nanospheres (NH2-MCM-41) were prepared via sol-gel route according to the literature [24] and surface grafting reaction. Typically, 1.69 g CTAB was added to aqueous solution of 800 mL deionized water and 0.46 g NaOH at 80 °C with vigorous stirring. After complete dissolution of CTAB, 7.4 g TEOS was dropped into the obtained solution slowly, followed by 2 h reaction and afterwards fast addition of 1.14 g KH550. After 1 h reaction, the product was obtained using the settlement, filtration and 5-time washing with ethanol. Subsequently, the products underwent drying process at 100 °C overnight. The schematic illustration of preparation of amino grafting MCM-41 was showed in Fig. 1 (a). 2.3. Preparation of GO nanosheets GO nanosheets were prepared according to the modified Hummer's Method [25]. In detail, 3 g graphite and 1.5 g NaNO3 were charged into 69 mL condensed H2SO4 at 0 °C using cooling system. Subsequently, 9 g KMnO4 was added to the suspension in 3 times under strong stirring below 20 °C, after which, the temperature was increased to 35 °C for oxidation process of 5 h. Afterwards, 138 mL deionized water was added to the suspension very slowly, followed by increasing the temperature to 98 °C and keeping for 15min. Additional 420 mL water and 102

Composites Part B 138 (2018) 101–112

Z. Li et al.

Fig. 1. (a) Preparation of amino grafting MCM-41 nanospheres; (b) Preparation of GO decorated with MCM-41 nanospheres.

Fig. 2. (a) FTIR spectra of GO, NH2-MCM-41 and [email protected]; (b) Raman spectra of GO, [email protected] and (c) XRD patterns of GO, NH2-MCM-41 and [email protected]

Instrumented Charpy Impact Tester (Germany) at an impact speed of 2.93 m/s. The samples were not notched with 50 mm × 6 mm × 4 mm, conforming to the DIN53753 standard. The combustion behavior was evaluated by cone calorimeter tester (CCT, FTT). At least two samples were measured according to ISO 5660-1 with geometric dimension of 100 mm × 100 mm × 4 mm under heat flux of 50 kW/m2 to obtain average value.

conveying pipe. The constant weight of specimen (15 ± 0.2 mg) underwent the thermal degradation and the evolved volatiles were directed to FTIR chamber for spectra acquisition. DMA measurement was carried out using dynamic mechanical analyzer (Q800, TA Instruments) in the multi-frequency strain mode with the single cantilever. The temperature was increased from the room temperature to 270 °C at 3 °C/min with amplitude of 30 μm and frequency of 1 Hz. The unnotched Charpy impact behavior was measured using Zorn Standal 103

Composites Part B 138 (2018) 101–112

Z. Li et al.

Fig. 3. (a) XPS full spectra, (b) C1s spectra, (c) N1s spectra and (d) Si 2p spectra of [email protected]

3. Results and discussion

with N2 sorption measurement, the mesoporous MCM-41 possessed targeted pore property (specific area 106 m2/g and pore size 2.0 nm). In parallel, the peak at 9.1° was typically attributed to (001) of GO with lamellar spacing (0.97 nm). The investigation of the [email protected] illustrated that (110) from MCM-41 and (001) from GO were also visible in [email protected] The slight shift of (001) crystalline faces of [email protected] MCM-41 revealed the interaction of GO and MCM-41. In order to further confirm the reaction between GO and NH2-MCM41, XPS was performed and showed in Fig. 3. In Fig. 3(a), Si, N, O and C were present in [email protected], which were consistent with FTIR result. C1s spectra fitting analysis (Fig. 3(b)) showed the peaks presence of OC]O (288.2eV), C]O (287.0eV), C-N (285.4eV), C-OH (285.3eV), C]C (284.4eV) and C-C (284.9), which indicated the partial reduction of GO compared with the reported [28]. Meanwhile, N1s fitting investigation (Fig. 3(c)) demonstrated the respective linking of N atom to C]O (400.9eV), C]C (402.3eV) and C-C (402.9eV), which manifested that NH2-MCM-41 was successful grating on GO through reaction between amino group and epoxy and carboxyl groups. Si2p spectra analysis (Fig. 3(d)) revealed the predominant presence of Si-O-Si, together with secondary Si-C and Si-OH groups. SEM and TEM results were showed in Fig. 4. In Fig. 4 (a) and (b), GO nanosheets possessed enormous wrinkles and lamellar structure (composed of multi-sheet and single-sheet structure). In parallel, FIBSEM and TEM observation (Fig. 4 (c) and (d)) illustrated that MCM-41 was spherical or sphere-like with about 100 nm diameter, accompanied by mesopores of about 2 nm aligning from one top to another. In terms

3.1. Nanohybrid characterization FTIR spectra of GO, NH2-MCM-41 and [email protected] were showed in Fig. 2 (a). In terms of GO, C]O and C-O bonds were present, which indicated the oxidation state of GO. In parallel, the peaks at 1221 cm−1 and 1059 cm−1 were attributed to Si-O-Si stretching vibration of NH2MCM-41. Contrastively, [email protected] illustrated the combined peaks of GO and NH2-MCM-41. Interestingly, C-N emerged in [email protected] with the location shifted to higher wavenumber (1400 cm−1), which illuminated the chemical grafting reaction between NH2-MCM-41 and active groups (epoxy group and carboxyl group) of GO [27]. Raman spectra result was showed in Fig. 2 (b) with commercial graphene nanoplates as control sample. Notably, two peaks at 1380 cm−1 (D peak) and 1591 cm−1 (G peak) appeared, which were correlated to sp3 and sp2 hybridized carbon respectively [28]. GO exhibited blue shifted D and red shifted G peak, which demonstrated strong oxidation state of GO, together with significantly increased intensity ratio of D and G peak of GO (1.64) compared with graphene (1.35). Comparatively, D and G location of [email protected] were 1350 cm−1 and 1594 cm−1 between GO and graphene, which evidenced the partial reduction of GO, together with between-lied intensity ration ID/IG value (1.54). XRD analysis was shown in Fig. 2 (c). Three peaks at 2.1, 3.7 and 4.3° in MCM-41 were attributed to (100), (110) and (200) separately, which illustrated hexagonal alignment of mesopores. In combination 104

Composites Part B 138 (2018) 101–112

Z. Li et al.

Fig. 4. (a) SEM and (b) TEM image of GO nanosheets; (c) FIB-SEM and (d) TEM image of MCM-41 nanospheres; (e) SEM and (g) TEM image of [email protected] nanohybrid and (g) EDS result of [email protected]

Fig. 5. TEM images of (a) EP/MCM-41, (b) EP/GO, (c) EP/GOMCM-41, (d) and (f) EP/[email protected] and (e) HAADF image with Si mapping of EP/[email protected]

3.2. EP nanocomposite structural characterization

of [email protected], there was no presence of notable aggregated MCM-41 zones in GO nanosheets in SEM image (Fig. 4 (e)) with fluffy surfaces, which was distinct from GO nanosheets. TEM observation (Fig. 4 (f) and (h)) disclosed the coherent attachment of MCM-41 on GO surface to form assembly structure, leaving wrinkles between MCM-41 nanospheres. Even, some MCM-41 nanospheres were encapsulated by GO nanosheets. In order to acquire element information, EDS connected to SEM was performed showed in (Fig. 4 (g)). [email protected] was composed of three elements (C, O and Si) with silicon atomic percentage nearly 6% in detecting zone.

TEM and XRD analysis were employed to investigate the dispersion of nanofillers in EP (Fig. 5 and Fig. SI.1). In TEM images (Fig. 5 (a)), MCM-41 were dispersed in EP homogeneously with few aggregates. High-magnification images manifested that MCM-41 mesopores were preserved without damage under strong shear force (Fig. SI.2). In parallel, GO nanosheets was dispersed in EP in partially exfoliated state with few aggregates and enlarged lamellar spacing relative to pristine GO (Fig. 5 (b)). TEM observation (Fig. 5 (d)) of EP/[email protected] 105

Composites Part B 138 (2018) 101–112

Z. Li et al.

dispersion state of MCM-41 and GO without MCM-41 attached to GO. Finally, SAED results demonstrated no clear diffraction patterns, which was indicative of exfoliated or even disordered structure.

Table 1 TG and DTG data of EP nanocomposites at N2 atmosphere. Samples

EP EP/MCM-41 EP/GO EP/[email protected] EP/GO-MCM-41

N2 atmosphere T5wt% (°C)

Tmax (°C)

C700 (%)

Rm (%/°C)

381 378 373 362 389

410 410 410 402 420

17.7 17.2 18.4 26.1 16.1

1.7 1.5 1.7 1.2 1.9

3.3. Thermal degradation behaviour Thermal degradation results were revealed in Table 1, Fig. 6 (a) and Fig. 6 (b). The addition of MCM-41 (378 °C), GO (373 °C) and [email protected] MCM-41 (362 °C) uniformly decreased T5wt% compared with pristine EP (381 °C), which indicated that nanofillers contributed to initial thermal degradation of EP possibly due to abundant active groups on MCM-41 and GO [29]. Intriguingly, EP/GO-MCM-41 showed increased T5wt% compared with EP/[email protected] and EP, which was possibly due to synergistic reinforcement of GO and MCM-41 in terms of volatile retention [30,31]. In contrast, [email protected] weakened volatile retention effect due to the occupation of MCM-41 on GO surface. Similarly, Tmax showed the similar trend to T5wt%. Notably, the maximum weight derivation of EP/[email protected] was approximately 1.2%/°C, significantly lower than its counterparts, which revealed that [email protected] exerted more extensive effect in alleviating thermal degradation [32]. In terms of char yield at 700 °C, EP/[email protected] possessed about 26.1%, significantly higher than EP, EP/GO, EP/MCM-41 and EP/EP-MCM-41. Amongst, the comparison of EP/[email protected] and EP/GO-MCM-41 demonstrated that the spatially preferential assembly of GO and MCM41 exerted stronger charring capacity than that of direct addition of GO and MCM-41. In parallel, thermal-oxidation degradation results (Table SI.1 and Fig. SI.3) revealed significantly reduced 1st peak thermaloxidation rate and increased char yield between ca 550 °C and 650 °C.

T5wt%: Temperature at 5% weight loss Tmax: Temperature at maximum weight loss. C700: Char yield at 700 °C Rm: Maximum weight loss rate.

3.4. Combustion behaviour Cone calorimeter test (CCT) was regarded as the most appropriate to reflect realistic fire situation, which enabled to simulate the entire combustion process from ignition to flame out and record heat release, smoke release and gas release profiles [33,34]. CCT results were illustrated in Table 2 and Fig. 7. Herein, TTI values of EP nanocomposites were relatively lower than neat EP with EP/[email protected] occupying the lowest probably due to the decreased initial thermal stability. In HRR curves (Fig. 7 (a)), pHRR value of EP/[email protected] (588 ± 14 kW/m2) was magnificently reduced with respect to EP (879 ± 22 kW/m2), EP/GO (979 ± 18 kW/m2), EP/MCM-41 (939 ± 23 kW/m2) and EP/GO-MCM-41 (697 ± 20 kW/m2), which indicated that the combination of GO and MCM-41 yielded the remarkable synergistic effect. Herein, the isolated addition of GO and MCM-41 even increased pHRR value slightly due to the deterioration of carbonaceous layer (skeleton effect). Additionally, covalent assembly of MCM-41 and GO to form hierarchical nanohybrid had a better fire performance than the direct addition. In other word, the spatial relationship between GO and MCM-41 exerted critical influences in fire behavior. The curve comparison revealed that one obvious peak emerged in the late stage of combustion in EP/GO-MCM-41 but not in EP/[email protected], which indicated that EP/GO-MCM-41 yielded mechanically weaker char compared with EP/[email protected] Weaker char broke under the intensive impact of volatiles, which further resulted in

Fig. 6. (a) TG and (b) DTG curves of EP, EP/GO, EP/MCM-41 and EP/[email protected] at N2 atmosphere.

revealed that GO were also exfoliated homogenously with MCM-41 attached on GO surface firmly. The HAADF image with silicon mapping further evidenced the selective attachment of MCM-41 on GO nanosheets in EP nanocomposites (Fig. 5 (e)). Comparatively, TEM observation (Fig. 5 (c)) of EP/GO-MCM identically showed nanoTable 2 CCT data of EP and EP nanocomposites (50 kW/m2). TTI (s) EP EP/MCM-41 EP/GO EP/[email protected] EP/GO-MCM-41

67 60 56 53 63

± ± ± ± ±

1 2 1 1 1

pHHR (kW/m2)

THR (kJ/m2)

THR/TML (MJ/(m2 g))

FIGRA (kW/(m2.s))

Char yield (%)

879 939 979 588 697

108 108 107 101 104

2.39 2.48 2.47 2.45 2.32

5.04 6.75 6.35 3.46 3.60

13.2 15.5 10.4 16.3 15.3

± ± ± ± ±

22 23 18 14 20

± ± ± ± ±

3 5 4 4 3

± 0.01 + 0.01 ± 0.01 ± 0.02 ± 0.01

± ± ± ± ±

0.02 0.04 0.03 0.02 0.02

TTI: Time to ignition pHRR: Peak heat release rate THR: Total heat release THR/TML: Efficient combustion heat TML: Total mass loss FIGRA: Fire growth index.

106

± ± ± ± ±

0.2 0.3 0.3 0.1 0.1

Composites Part B 138 (2018) 101–112

Z. Li et al.

Fig. 7. (a) HRR, (b) Mass profiles, (c) SPR and (d) COP of EP, EP/MCM-41, EP/GO, EP/GO-MCM-41 and EP/[email protected]

fire growth. THR results revealed the similar behavior to pHRR. Fire growth rate index (FIGRA) investigation (pHRR divided by time to pHRR) illuminated that EP/[email protected] possessed the significantly reduced FIGRA value compared with its counterparts. Moreover, EP/ GO-MCM-41 showed slightly higher FIGRA value than EP/[email protected], which illustrated that the direct addition of GO and MCM-41 generated higher fire growth rate than spatial assembly [email protected] Mass profile study (Fig. 7 (b)) exhibited that char yield of EP/[email protected] MCM-41 was almost equal to those of EP/MCM-41 and EP/GO-MCM41, higher than EP and EP/GO, which was in agreement with HRR result in terms of reduced fire hazard compared with neat EP. Interestingly, EP/MCM-41 revealed deteriorated pHRR but retained high char yield with mass loss focused in the early stage, which was probably due to the catalytic charring of MCM-41 and resultant damage to EP char quality. In addition, smoke production rate (SPR) profile (Fig. 7 (c)) showed that EP/[email protected] possessed significantly peak SPR with respective reduction rate of 32%, 39%, 32% and 10% in contrast to EP, EP/GO, EP/MCM-41 and EP/GO-MCM-41, which illuminated the notable impact of spatial relationship (neighboring synergy) of GO and MCM-41 on smoke production behavior. In parallel, the peak CO production rate of EP/[email protected] was also remarkably reduced. Reasonably, COP and smoke profiles were also two key parameters determining fire retardancy in combination with heat profile. In conclusion, [email protected] favoured fire retardancy compared with GO, MCM-41 and GO-MCM-41. Meanwhile, spatial assembly of MCM-41 and GO imparted EP with better fire retardancy than random dispersion disclosed in Fig. 5.

exhibited in Fig. 8 and Fig. SI.4 and Fig. SI.5. From horizontal and top view, EP yielded the char with loose and isolated fragments with enormous pores in interior layer and isolated fragments in exterior layer. The addition of MCM-41 and GO deteriorated loose and isolated structure, together with more fractured pores in interior layer and heterogamous rod-like components emerging in exterior layer. In contrast, the addition of [email protected] produced char with compact and integral structure and larger dimension. SEM observation illustrated that interior and exterior layers possessed the char morphology without notable pores and fragments, which significantly inhibited the transmission of heat, oxygen and heat between underlying matrix and flaming zone. Reasonably, the strong char enabled to resist breakage and fracture from the volatiles impact due to its mechanical robustness during combustion (Fig. 7 (a)). The direct addition of GO and MCM-41 resulted in the char quality lied between EP and EP/[email protected] with isolated chars connected with big pores. EDS result comparison (Fig. SI.5) demonstrated that enormous silicon was present in exterior layer of EP/GO-MCM-41 but not EP/[email protected], which indicated that the migration of Si element was hindered by GO nanosheets due to the spatial relationship. Raman spectra analysis was performed to understand char composition and structure (Fig. 9 (a)). In terms of Raman spectra, D and G band at 1357 cm−1 and 1586 cm−1 emerged, which were assigned to A1g breathing vibration of sp3 and E2g in-plane stretching vibration of sp2 six-ring carbon pairs respectively [28]. The intensity ratio of D and G band (ID/IG) was employed to evaluate micro-crystalline size, with the lower ID/IG indicative of higher micro-crystalline size and graphitization. Herein, the char from EP/[email protected] possessed ID/IG value of 2.63, slightly lower than those of EP/MCM-41, EP/GO and EP, which indicated that EP/[email protected] yielded the char with higher microcrystalline size. The higher micro-crystalline size laid the foundation for

3.5. Carbonaceous layer analysis The digital images and SEM analysis of the char after CCT were 107

Composites Part B 138 (2018) 101–112

Z. Li et al.

Fig. 8. Digital images of chars from horizontal view and top view of (a1, a2) EP, (b1,b2) EP/GOMCM-41 and (c1,c2) EP/[email protected]; SEM observation of interior and exterior morphology of (a3, a4) EP, (b3,b4) EP/GO-MCM-41 and (c3,c4) EP/[email protected]

order to better analyze volatile profiles, the selected components were illuminated in detail (Fig. 10). The incorporation of GO and MCM-41 did not decrease release intensity of aliphatic compounds, aromatic compounds and carbonyl compounds. Contrastively, [email protected] imparted EP with significantly reduced production of the respective compounds. Notably, NH3 release intensity of EP/[email protected] was magnificently higher than its counterparts, which was possibly due to the increased catalytic elimination capacity. In theory, the reduced release of volatiles favoured the suppression of combustion process, together with the enhanced release of NH3 due to the dilution effect. Based on the above analysis, the probable fire-retardant mechanism of EP/[email protected] was put forward in Fig. 11. In vapour phase, the incorporation of [email protected] into EP increased NH3 release, which diluted flammable gases and suppressed fire growth. In condensed phase, the intensively protective chars (SEM and digital images) were generated via incorporation of the assembly of MCM-41 and GO. In detail, the isolated incorporation of GO and MCM-41 damaged char structure of EP due to good charring capacity of EP itself (Fig. 8). In contrast, the spatially preferential assembly of GO and MCM-41 remarkably strengthened the char structure and released more nonflammable gases (Fig. 10 (a)), resulting in improved fire retardancy. Specially, MCM-41 tended to promote charring owing to Lewis acid sites Fig. 7 (b). The newly-formed char covered GO nanosheets due to the neighbouring location, further favouring char quality (Raman spectra, SEM and digital image) and avoiding thermal degradation of GO nanosheets [28]. In the sense, the combustion process was weakened, accompanied by the generation of less volatile (Fig. 10). Comparatively, direct addition of GO and MCM-41 yielded relatively lowerquality char, contributing to fire retardancy partially. Comprehensively, the fire retardancy of EP was utmost improved based on spatially preferential combination of GO and MCM-41 via neighboring synergy mechanism.

mechanical robust char with strong stiffness [35]. 3.6. Degradation behavior in condensed phase vt FTIR of EP and EP/[email protected] at various times were exhibited in Fig. 9 (b) and (c). In terms of EP, the peaks at 3037 cm−1 (stretching vibration mode of C]C-H bond) became more intensive compared with that of C-H located between 2980 cm−1 and 1840 cm−1. In addition, the peaks at 880 cm−1(C]C-H), 1368 cm−1 (C-H) and 1600 cm−1 (C]C) also showed the enhanced peak intensity, which indicated the formation of aromatic compounds. In contrast, the peaks located at 1500 cm−1 (benzene ring), 1100 cm−1 (ether) and 817 cm−1 (substituted benzene) became relatively weaker, which indicated that the substituted benzene and ether groups were released into volatiles. In parallel, EP/[email protected] followed the identical thermal degradation route in condensed phase to EP with formation of char with enormous aromatic compounds. Meanwhile, C]C bond (1600 cm−1) of EP/[email protected] MCM-41 appeared at ca 30min while EP at ca 35min, which implied the initial degradation of EP and showed good agreement with TG result. The earlier formation of C]C containing compounds generally favored the protection behavior of the carbonaceous layers towards the underlying matrix. 3.7. Volatile evolution analysis TG-FTIR results were demonstrated in Fig. SI.6 and Fig. 10. TG-FTIR spectra at maximum volatile release rate illustrated that various nanofillers did not change volatile species but release intensity. The main released components included aromatic compounds (1510 cm−1), aliphatic compounds (2980 cm−1), carbonyl compound (1810 cm−1), ammonia gas (3400 cm−1), carbon monoxide (2180 cm−1), carbon dioxide (2360 cm−1). In parallel, EP/[email protected] imparted EP with remarkably lower release intensity than EP/MCM-41 and EP/GO. In 108

Composites Part B 138 (2018) 101–112

Z. Li et al.

Fig. 9. (a) Raman spectra of char from EP and EP nanocomposites; vt FTIR spectra of (b) EP and (c) EP/[email protected]

much rougher fracture surface with enormous wrinkles, which was probably due to plasticizing effect of MCM-41 (Fig. 12 (b)). Notably, EP/[email protected] manifested more tiny plastic deformation areas with more wrinkles (Fig. 13 (e)), which ascertained more consumed energy and higher toughness. Finally, the morphology of EP/GO-MCM-41 was similar to EP with limited numbers of plastic deformation zones. The direct addition of GO and MCM-41 gave rise to EP nanocomposites with randomly dispersed nanofillers. The cracks were ape to pass MCM-41 nanospheres and GO nanosheets in propagation. The spatially preferential assembly of MCM-41 on GO synergistically acted as effective barriers to cracks with GO consuming the energy through ¨pining¨ effect and ¨bridge¨ effect and MCM-41 plasticizing EP matrix.

3.8. Mechanical property investigation 3.8.1. Dynamic mechanical property Dynamic mechanical property was showed in Fig. 12 (a) and (b). The incorporation of GO, MCM-41 and GO-MCM-41 slightly decreased storage modulus (G′) compared with neat EP, while [email protected] increased G′ notably above 50 °C probably due to promoted the dispersion of MCM-41 via spatially preferential arrangement. In parallel, the investigation of tan δ illuminated that incorporation of various nanofillers decreased glass transition temperature (Tg) of EP, accompanied by the most reduction of EP/MCM-41 possibly due to the damage to EP network [36] and/or the plasticizing effect. Tan δ value of EP and EP nanocomposites nearly remained invariable, which illustrated that the internal fraction of EP segments did not change with addition of nanofillers.

4. Conclusions In the manuscript, the hierarchical nanohybrid ([email protected]) of MCM-41 nanospheres spatially preferential location on GO nanosheets was prepared successfully on basis of covalent assembly. The nanohybrid was confirmed by FTIR, Raman spectra, XRD, XPS, SEM and TEM. 2 wt% [email protected] showed the exfoliated nano-dispersion in EP matrix, evidenced by TEM and XRD. In thermal degradation process, EP/[email protected] demonstrated notably improved char yield, accompanied by significantly decreased peak degradation rate. Meanwhile, a 40.0% reduction of pHRR in CCT was observed in EP/[email protected] compared with EP/GO, together with remarkable reduction of peak smoke production and the formation of high-quality char. Furthermore, predominant volatiles were decreased with exception of NH3. Based on detailed analysis, the neighboring synergy associated with GO

3.8.2. Impact behavior In Fig. 12 (c), the addition of GO, MCM-41, [email protected] and GOMCM-41 increased unnotched impact strength of EP (8.2 ± 3.5 kJ/m2) to 11.8 ± 3.0 kJ/m2, 17.2 ± 2.8 kJ/m2, 13.4 ± 2.4 kJ/m2, 7.2 ± 2.8 kJ/m2 correspondingly, which manifested the most enhancement of [email protected] [37]. The mechanism of enhanced impact strength was investigated by the analysis of fracture surface (Fig. 13). In terms of EP (Fig. 13 (a)), the crack propagated smoothly in approximately straight line nearly without notable resistance in walking path. In terms of EP/GO, slightly more crack aggregates happened, which manifested that GO nanosheets were slightly beneficial in improving fracture toughness. In comparison, MCM-41 incorporation generated 109

Composites Part B 138 (2018) 101–112

Z. Li et al.

Fig. 10. (a) Aromatic compounds, (b) aliphatic ether, (c) carbonyl compounds and (d) ammonia gas evolution profile of EP, EP/GO, EP/MCM-41 and EP/[email protected] Fig. 11. Proposed fire retardancy mechanism of EP/[email protected] MCM-41

110

Composites Part B 138 (2018) 101–112

Z. Li et al.

Fig. 12. (a) Storage modulus curves; (b) tan δ curves and (c) unnotched impact behavior of EP and EP nanocomposites. Fig. 13. Fracture surface morphology of (a) EP, (b) EP/GO, (c) EP/MCM-41, (d) EP/GO-MCM-41 and (e) EP/[email protected]; High magnification images of (f) EP/GO-MCM-41 and (g) EP/[email protected] MCM-41.

111

Composites Part B 138 (2018) 101–112

Z. Li et al.

nanosheets wrapped by charring due to MCM-41 catalysis was proposed to account for improved fire retardancy, together with NH3 dilution effect. In parallel, EP/[email protected] possessed higher impact strength (13.4 ± 2.4 kJ/m2) with respect to EP (8.2 ± 3.5 kJ/m2). Moreover, the hierarchical nanohybrid with spatially preferential assembly of MCM-41 on GO imparted EP with higher impact value than direct addition, which was involved in synergistic effect in hindering crack propagation.

2016;4(6):2147–57. [14] Zammarano M, Franceschi M, Bellayer S, Gilman JW, Meriani S. Preparation and flame resistance properties of revolutionary self-extinguishing epoxy nanocomposites based on layered double hydroxides. Polymer 2005;46(22):9314–28. [15] Guan F-L, Gui C-X, Zhang H-B, Jiang Z-G, Jiang Y, Yu Z-Z. Enhanced thermal conductivity and satisfactory flame retardancy of epoxy/alumina composites by combination with graphene nanoplatelets and magnesium hydroxide. Compos Part B Eng 2016;98:134–40. [16] Li Z, Wang D-Y. Nano-architectured mesoporous silica decorated with ultrafine Co3O4 toward an efficient way to delaying ignition and improving fire retardancy of polystyrene. Mater Des 2017;129:69–81. [17] Dreyer DR, Park S, Bielawski CW, Ruoff RS. The chemistry of graphene oxide. Chem Soc Rev 2010;39(1):228–40. [18] Wang Z, Wei P, Qian Y, Liu J. The synthesis of a novel graphene-based inorganic–organic hybrid flame retardant and its application in epoxy resin. Compos Part B: Eng 2014;60:341–9. [19] Wang R, Zhuo D, Weng Z, Wu L, Cheng X, Zhou Y, et al. A novel nanosilica/graphene oxide hybrid and its flame retarding epoxy resin with simultaneously improved mechanical, thermal conductivity, and dielectric properties. J Maters Chem A 2015;3(18):9826–36. [20] Wang D, Zhou K, Yang W, Xing W, Hu Y, Gong X. Surface modification of graphene with layered molybdenum disulfide and their synergistic reinforcement on reducing fire hazards of epoxy resins. Ind Eng Chem Res 2013;52(50):17882–90. [21] Vallet-Regi M, Ramila A, Del Real R, Pérez-Pariente J. A new property of MCM-41: drug delivery system. Chem Mater 2001;13(2):308–11. [22] Wang N, Zhang J, Fang Q, Hui D. Influence of mesoporous fillers with PP-g-MA on flammability and tensile behavior of polypropylene composites. Compos Part B Eng 2013;44(1):467–71. [23] Wang N, Mi L, Wu Y, Zhang J, Fang Q. Double-layered co-microencapsulated ammonium polyphosphate and mesoporous MCM-41 in intumescent flame-retardant natural rubber composites. J Therm Anal Calorim 2014;115(2):1173–81. [24] Cai Q, Luo Z-S, Pang W-Q, Fan Y-W, Chen X-H, Cui F-Z. Dilute solution routes to various controllable morphologies of MCM-41 silica with a basic medium. Chem Mater 2001;13(2):258–63. [25] Hummers Jr. WS, Offeman RE. Preparation of graphitic oxide. J Am Chem Soc 1958;80(6):1339. [26] Zhang H, Tong C, Sha J, Liu B, Lü C. Fluorescent mesoporous silica nanoparticles functionalized graphene oxide: a facile FRET-based ratiometric probe for Hg2+. Sensors Actuators B Chem 2015;206:181–9. [27] Qian Y, Wei P, Jiang P, Zhao X, Yu H. Synthesis of a novel hybrid synergistic flame retardant and its application in PP/IFR. Polym Degrad Stabil 2011;96(6):1134–40. [28] Guo Y, Bao C, Song L, Yuan B, Hu Y. In situ polymerization of graphene, graphite oxide, and functionalized graphite oxide into epoxy resin and comparison study of on-the-flame behavior. Ind Eng Chem Res 2011;50(13):7772–83. [29] Wang D-Y, Das A, Leuteritz A, Boldt R, Häußler L, Wagenknecht U, et al. Thermal degradation behaviors of a novel nanocomposite based on polypropylene and Co-Al layered double hydroxide. Polym Degrad Stabil 2011;96(3):285–90. [30] Ran S, Chen C, Guo Z, Fang Z. Char barrier effect of graphene nanoplatelets on the flame retardancy and thermal stability of high-density polyethylene flame-retarded by brominated polystyrene. J Appl Polym Sci 2014(15):131. [31] Zhang F-A, Lee D-K, Pinnavaia TJ. PMMA/mesoporous silica nanocomposites: effect of framework structure and pore size on thermomechanical properties. Polym Chem 2010;1(1):107–13. [32] Zhao X, Yang L, Martin FH, Zhang X-Q, Wang R, Wang D-Y. Influence of phenylphosphonate based flame retardant on epoxy/glass fiber reinforced composites (GRE): Flammability, mechanical and thermal stability properties. Compos Part B Eng 2017;110:511–9. [33] Schartel B, Hull TR. Development of fire retarded materials—Interpretation of cone calorimeter data. Fire Mater 2007;31(5):327–54. [34] Pan Y-T, Wang X, Li Z, Wang D-Y. A facile approach towards large-scale synthesis of hierarchically nanoporous [email protected] 0D/1D hybrid and its effect on flammability, thermal stability and mechanical property of flexible poly (vinyl chloride). Compos Part B Eng 2017;110:46–55. [35] Zhang J, Kong Q, Yang L, Wang D-Y. Few layered Co(OH)2 ultrathin nanosheetbased polyurethane nanocomposites with reduced fire hazard: from eco-friendly flame retardance to sustainable recycling. Green Chem 2016;18(10):3066–74. [36] Ye Y, Chen H, Wu J, Ye L. High impact strength epoxy nanocomposites with natural nanotubes. Polymer 2007;48(21):6426–33. [37] Kang W-S, Rhee KY, Park S-J. Influence of surface energetics of graphene oxide on fracture toughness of epoxy nanocomposites. Compos Part B Eng 2017;114:175–83.

Acknowledgements This research is funded by Spanish Ministry of Economy and Competitiveness (MINECO) under Ramón y Cajal fellowship (RYC2012-10737) and COST Action CM1302 (Smart Inorganic Polymers). Likewise, the author was grateful to Dr. Miguel Castillo and Dr. Juan Pedro Fernández in IMDEA Materials Institute for helpful discussion. Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx. doi.org/10.1016/j.compositesb.2017.11.001. References [1] Liu S, Yan H, Fang Z, Wang H. Effect of graphene nanosheets on morphology, thermal stability and flame retardancy of epoxy resin. Compos Sci Technol 2014;90:40–7. [2] Li Z, Liu L, González AJ, Wang D-Y. Bioinspired polydopamine induced assembly of ultrafine Fe (OH)3 nanoparticles on halloysite toward highly efficient fire retardancy of epoxy resin via an action of interfacial catalysis. Polym Chem 2017;8:3926–36. [3] Guo W, Yu B, Yuan Y, Song L, Hu Y. In situ preparation of reduced graphene oxide/ DOPO-based phosphonamidate hybrids towards high-performance epoxy nanocomposites. Compos Part B Eng 2017;123:154–64. [4] Zhang Q, Tian M, Wu Y, Lin G, Zhang L. Effect of particle size on the properties of Mg(OH)2-filled rubber composites. J Appl Polym Sci 2004;94(6):2341–6. [5] Dittrich B, Wartig KA, Hofmann D, Mülhaupt R, Schartel B. The influence of layered, spherical, and tubular carbon nanomaterials' concentration on the flame retardancy of polypropylene. Polym Compos 2015;36(7):1230–41. [6] Gallo E, Schartel B, Braun U, Russo P, Acierno D. Fire retardant synergisms between nanometric Fe2O3 and aluminum phosphinate in poly(butylene terephthalate). Polym Advan Technol 2011;22(12):2382–91. [7] Liu Y, Babu HV, Zhao J, Goñi-Urtiaga A, Sainz R, Ferritto R, et al. Effect of Cu-doped graphene on the flammability and thermal properties of epoxy composites. Compos Part B Eng 2016;89:108–16. [8] Jiang S-D, Bai Z-M, Tang G, Song L, Stec AA, Hull TR, et al. Synthesis of mesoporous [email protected] Co-Al layered double hydroxide spheres: layer-by-layer Method and their effects on the flame retardancy of epoxy resins. ACS Appl Mater Interfaces 2014;6(16):14076–86. [9] Gu J, Liang C, Zhao X, Gan B, Qiu H, Guo Y, et al. Highly thermally conductive flame-retardant epoxy nanocomposites with reduced ignitability and excellent electrical conductivities. Compos Sci Technol 2017;139:83–9. [10] Khalili P, Tshai K, Hui D, Kong I. Synergistic of ammonium polyphosphate and alumina trihydrate as fire retardants for natural fiber reinforced epoxy composite. Compos Part B Eng 2017;114:101–10. [11] Matykiewicz D, Przybyszewski B, Stanik R, Czulak A. Modification of glass reinforced epoxy composites by ammonium polyphosphate (APP) and melamine polyphosphate (PNA) during the resin powder molding process. Compos Part B Eng 2017;108:224–31. [12] Wang X, Xing W, Feng X, Yu B, Lu H, Song L, et al. The effect of metal oxide decorated graphene hybrids on the improved thermal stability and the reduced smoke toxicity in epoxy resins. Chem Eng J 2014;250:214–21. [13] Kalali EN, Wang X, Wang D-Y. Multifunctional intercalation in layered double hydroxide: toward multifunctional nanohybrids for epoxy resin. J Mater Chem A

112