Huihuan Wanga,
Chuan Lib,
Zhuang Houa,
Bolun Lia and
Haopeng Cai*ac
aSchool of Materials Science and Engineering, Wuhan University of Technology, Wuhan 430070, China
bShanghai Composites Science & Technology Co., Ltd, Shanghai 201112, China
cInstitute of Advanced Materials Manufacturing Equipment and Technology, Wuhan University of Technology, Wuhan 430070, China
First published on 1st March 2022
In order to develop a liquid oxygen-compatible (LOX-compatible) matrix resins for polymer-based fiber-reinforced composites, a novel phosphorus-containing imidazole derivative called VAD containing multifunctional groups was synthesized and used as a co-curing agent for epoxy resin (EP) with simultaneous LOX-compatibility and mechanical improvement. A phosphorus group was introduced into the EP to capture the free radicals generated during the pyrolysis of the polymer to improve LOX compatibility, and the trimethylene group was introduced as a flexible spacer to enhance the toughness of the cured material. In comparison to pure EP, the modified EP with only 2.5 wt% VAD showed excellent mechanical properties with 23.0% and 75.6% increase in tensile and impact strength, respectively. Furthermore, as the content of VAD increased, a thermoset compatible with LOX (according to the liquid oxygen impact test) was obtained, and the flame retardancy was improved (according to the limiting oxygen index test). However, there was no significant sacrifice of transparency or thermal stability. In addition, the LOX compatibility mechanism was analyzed using X-ray photoelectron spectroscopy. As an efficient multi-functional modifier, VAD has a bright future in the modification realm of EP materials.
According to the hot spot theory, when the polymer material is impacted, the mechanical energy in certain local regions (hot spots) is converted to thermal energy to cause local high temperatures, generate free radicals, initiate the chain reaction between polymer and oxygen, and lead to liquid oxygen shock sensitivity.8 Fundamentally, the incompatible reaction between EP and LOX is essentially consistent with its oxidation and combustion in oxygen. Therefore, the LOX compatibility of EP can be improved by improving its thermal stability and flame retardancy.9
Numerous phosphorus-based flame retardants have been utilized for exploiting flame-retardant EPs because of their superior flame-retardant properties, environmental safety, low toxicity, and non-halogen nature. Particularly, 9,10-dihydro-9-oxa-10-phosphaphenanthrene-10-oxide (DOPO) has been widely used in the field of flame retardant because of its strong molecular designability and high reaction activity.10–12 Given this, DOPO and its derivatives can also enhance the LOX compatibility of EP. LI et al. synthesized a new type of epoxy hybrid containing phosphorus/silicon from an epoxy resin containing DOPO and 3-glycidoxy-propyltrimethoxysilane (GLYMO), which achieved compatibility with LOX and possessed significantly enhanced thermal stability.13 Wu et al. utilized DOPO to chemically modify bisphenol A and bisphenol F EP, the LOX compatibility of the modified resin was significantly enhanced, better thermal stability and flame retardancy were also observed.14
The introduction of phosphorus-based flame retardants into the EP system can improve its compatibility with LOX. Besides, due to the inferior properties of the phosphorus-containing compounds and the physical blending of the unreacted flame retardant, the other original properties of the EP system deteriorate sharply, especially Tg and mechanical properties.15,16 The phosphorus flame retardant can achieve an excellent flame retardant effect only when it was grafted onto the cross-linked network.17,18 Therefore, it is instructive for academic and industrial fields to resolve the contradiction of improving LOX compatibility while maintaining or even improving the mechanical properties of EP by adding a single compound.
The methods to improve the toughness and strength of EP mainly include flexible polymer toughening and nano-material toughening.19–22 In most cases, EP toughened by nano-materials is prone to phase separation, which reduces the high strength of the matrix and often sacrifices thermal stability, which increases the difficulty of processing.23,24 The stress relaxation at the crack tip or the strength of the molecular chain is thought to control the fracture toughness of EP at ultra-low temperatures. The degree of cross-linking should be increased to increase the strength of the molecular chain, but the elasticity of the resin will be lost, reducing fracture toughness. For this reason, the epoxy hybrid containing multiple functional groups can be selected, with the rigid chain carrying the load and the flexible chain alleviating stress concentration.
In this work, a phosphorus-containing imidazole derivative VAD whose imidazole ring and phosphophenanthrene group were linked by an aminopropyl chain was designed and synthesized; it can be used as a multifunctional reactive modifier to form a block copolymer with EP. By adjusting the microstructure of EP, flexible segments were introduced in the three-dimensional cross-linking network to improve its molecular flexibility. As a reactive modifier, VAD caused no phase separation during the curing reaction. Whereas, the sacrifice of mechanical properties to increase LOX compatibility was absent. Architecturally, VAD contains a variety of functional groups, phosphophenanthrene as a rigid unit load, long-chain methylene as flexible chain retardation and stress concentration, and a DOPO drape as a flame retardant group. For the purpose of investigating the impact of VAD on the properties of EP matrix, thermal stability, the LOX compatibility, flame retardancy, and mechanical properties of pure and modified EPs were studied and analysed in detail.
FTIR spectra of VA, AI, DOPO and VAD are shown in Fig. 1. The stretching vibration absorption peaks of –OH and –CH3 at 3054 cm−1 and 2962 cm−1 are derived from VA, the stretching vibration absorption peaks of –CH2 at 2837 cm−1 and 2740 cm−1 are derived from AI as well as the characteristic peak of imidazole appearing at 1519 cm−1 were observed,25 while the typical absorption peak of –CHO at 1665 cm−1 disappeared in the FTIR spectrum of VAD, indicating the complete reaction between VA and AI. Furthermore, the double peaks of –NH2 at 3358 cm−1 and 3296 cm−1 converted to the single peak of –NH at 3399 cm−1, as well as the characteristic peak of P–H at 2436 cm−1 from DOPO disappeared, illustrating the nucleophilic addition reaction between the CN double bond and P–H was substantially completed. Additionally, the absorption peak at 1601, 856, and 756 cm−1 are attributed to the benzene ring, and the absorption peaks at 1282 cm−1, 1087 cm−1, 1032 cm−1 and 921 cm−1 are assigned to the stretching vibrations of P
O, C–N, C–O–C, and P–O–C, respectively. The chemical structure of VAD was validated by the information provided above.
Fig. 2 presents 1H NMR and 31P NMR spectra of VAD. As depicted in Fig. 2(a), a–f corresponds to six types of hydrogen (e, g and h were the same type). 8.7–9.3 ppm belonged to the reactive hydrogen chemical shift of –OH; the chemical shifts of protons, hydrogens of benzene and imidazole groups ranged from 6.7 to 8.2 ppm; 6.6 ppm corresponding to the chemical shift of N–H; 3.8 ppm, 3.7 ppm, 2.3–2.4 ppm and 1.6–1.7 ppm, and were assigned to the He, Hf, Hg and Hh of the alkyl chain, respectively; 3.9–4 ppm was attributed to Hd. Furthermore, the proton peaks' integral area ratio was Ha:
Hb
:
Hc
:
Hd
:
H(e + g + h): Hf = 0.9
:
14.2
:
1
:
1.1
:
6
:
3.2, which agreed well with the theoretical value. Two peaks at 30.52 and 33.09 ppm were observed as shown in Fig. 2(b), which were ascribed to the steric hindrance effect of the phosphaphenanthrene group.26 The results further validated the chemical structure of VAD.
![]() | ||
Fig. 3 TEM images of pure EP (a) and EP/VAD-5 (b). (c) P-mapping of EP/VAD-7.5. (d) The reaction mechanism between VAD and EP. (e) Photograph of EP/VAD-x. |
Samples | T5% (°C) | Tmax (°C) | CY800 °C (%) |
---|---|---|---|
EP | 359.8 | 387.9 | 15.9 |
EP/VAD-2.5 | 347.9 | 381.1 | 18.2 |
EP/VAD-5 | 341.1 | 378.3 | 20.2 |
EP/VAD-7.5 | 337.9 | 374.1 | 23.6 |
![]() | (1) |
Sample | Liquid oxygen incompatibility | IRS (%) | LOI (%) | |||
---|---|---|---|---|---|---|
Burning | Explosion | Flash | Charring | |||
EP | 0 | 2 | 6 | 0 | 27 | 25.6 |
EP/VAD-2.5 | 0 | 1 | 0 | 4 | 12.5 | 29.5 |
EP/VAD-5 | 0 | 0 | 1 | 1 | 5 | 30.6 |
EP/VAD-7.5 | 0 | 0 | 0 | 0 | 0 | 31.2 |
As the content of VAD increased, the IRS of EP/VAD thermosets showed a downward tendency from 27% to 0% while the LOI value had an obvious increase from 25.6% to 31.2% as presented in Table 2, indicating that flame retardant element P can improve the LOX compatibility of EP. Moreover, during the 20 times of impact tests, there was no reaction detected for EP/VAD-7.5, demonstrating that EP/VAD-7.5 were compatible with LOX.
Combined with the above thermal stability analysis, the LOX impact test and the limiting oxygen index test results, the corresponding LOX compatibility mechanism is shown in Fig. 6. When EP sample was impacted by high energy in LOX, the mechanical energy was converted into thermal energy to decompose the hot spot area of the resin, and the H·, O·, and OH· high-energy free radicals were generated, which caused chain decomposition of the resin and lead to incompatible reactions.30 However, for EP modified by VAD, a large number of PO· and HPO2· free radicals were released during thermal degradation can capture these high energy free radicals, thus inhibiting the impact-sensitive reaction between EP and LOX.
By observing the micro-morphology of the EP/VAD-5 sample after the LOX impact test, a thin deposited carbon layer was observed on the surface of the failed sample (see Fig. S2†), which inhibited the further combustion of the polymer and prevented the occurrence of LOX incompatibility. Further analysis was shown in Chapter 2.6.
Specimen | The impact test | Element content (%) | ||||
---|---|---|---|---|---|---|
P | C | N | O | O/C | ||
EP | Before | — | 75.43 | 8.29 | 16.28 | 0.2158 |
After (no reaction) | — | 75.9 | 8.52 | 15.58 | 0.2053 | |
After (explosion) | — | 76.8 | 8.84 | 14.36 | 0.1869 | |
EP/VAD-5 | Before | 0.36 | 79.75 | 3.93 | 15.96 | 0.2001 |
After (no reaction) | 0.4 | 78.93 | 3.55 | 17.11 | 0.2168 | |
After (flash) | 0.87 | 78.05 | 4.59 | 16.49 | 0.2113 |
Generally, during the pyrolysis under the O2 atmosphere, EPs are oxidized firstly and subsequently carbonized. The oxidation denotes an increased O/C ratio, while the carbonization suggests decreased O/C ratio. The reaction between EPs and LOX during the impact test can be explained by this mechanism.13 As presented in Table 3, after the impact test, the O/C ratio of pure EP decreased, indicating that it was carbonized during the impact test; the increased O/C ratio of EP/VAD-5 indicated that oxidation appeared during the impact test. It revealed that the neat EP had poor LOX compatibility, while EP/VAD-5 showed better LOX compatibility. This behavior was consistent with the results that EP samples were impacted in LOX.
The P2p spectra of EP/VAD-5 before and after the impact test are presented in Fig. 7. The spectrum before the impact test consisted of P–C and OP–O bonds and was decided from the primary chemical structure of EP/VAD. Compared with samples before the impact test, the P2p peak of the unreacted samples after the impact test was slightly wider, indicating the formation of PO3 groups; the P2p peak of the samples with the reaction after the impact test shifted significantly to higher binding energy, and its strength raised obviously, revealing that the phosphorus-containing groups in EP/VAD samples were heated to decompose into a large number of PO3 groups existing in the form of polymetaphosphoric acid to prevent further combustion of EP. The above results showed that the flame retardant mechanism of organophosphorus compounds in the condensed phase was that the generated polymetaphosphoric acid facilitated the formation of compact and stable phosphorus-rich char layer, which hindered availably of the diffusion of oxygen into the combustion and prevented the polymer from further combustion, as shown in Fig. 6.
Table 3 also shows that P and O contents of samples after combustion are significantly higher than those before impact, which further proved the above conclusion.
![]() | (2) |
![]() | (3) |
Sample | Elastic modulus (MPa) | Tensile strength (MPa) | Flexural modulus (MPa) | Flexural strength (MPa) | Elongation at break (%) | KIC (MPa m−1/2) | Impact strength (KJ m−2) |
---|---|---|---|---|---|---|---|
EP | 2378.1 ± 45 | 65.6 ± 1.5 | 2947.8 ± 9.6 | 93.6 ± 0.1 | 3.66 ± 0.05 | 1.97 ± 0.12 | 18.39 ± 1.2 |
EP/VAD-2.5 | 2957.7 ± 51 | 80.7 ± 1.3 | 2963.0 ± 27.2 | 94.5 ± 0.5 | 6.38 ± 0.04 | 3.52 ± 0.09 | 30.51 ± 0.9 |
EP/VAD-5 | 2971.6 ± 44 | 70.1 ± 1.9 | 3060.7 ± 32.1 | 100.7 ± 0.7 | 3.75 ± 0.05 | 2.82 ± 0.13 | 25.78 ± 1.0 |
EP/VAD-7.5 | 3150.0 ± 53 | 66.0 ± 1.7 | 3214.6 ± 32.6 | 108.3 ± 0.7 | 3.11 ± 0.06 | 2.31 ± 0.15 | 19.91 ± 1.4 |
The addition of a small amount of VAD to the matrix can improve both strength and toughness. As shown in Table 4 and Fig. 8, EP/VAD samples exhibited greater tensile strength (80.7–66 MPa) and elastic modulus (2957.7–3150 MPa), compared with those of pure EP (65.5 MPa and 2378.1 MPa). Similar phenomena were detected in flexural strength and modulus. The improvement in tensile and flexural properties can be attributed to (1) the addition of VAD and the introduction of a large number of rigid benzene rings in the EP matrix, and (2) many hydroxyl groups in the epoxy main chain can form hydrogen bonds with imidazole rings, enhancing the intermolecular interaction.31
Furthermore, the impact strength, elongation at break and KIC of EP-VAD-2.5 increased greatly, reaching the highest value of 30.51 kJ m−2, 6.38% and 3.52 MPa m−1/2, respectively, and then decreased linearly with the increase of VAD content, which attributed to internal defects in the homogeneous system, and lead to the early failure of the material.
The specific toughening mode of VAD on EP thermosets can be explained according to the morphology of the fracture surface examined by SEM, as presented in Fig. 9. As a result of the lack of energy dissipation, pure EP showed a characteristic brittle section with a “radial” pattern indicating that the crack propagates rapidly as shown in Fig. 9(a). Whereas EP/VAD samples (Fig. 9(b) and (c)) manifested the “dimple-like” ductile fracture characteristics,19 a significant increase in ripples and folds was observed, certifying that the crack deflection inhibited the rapid fracture of the sample. The relatively large number of cracks and rougher fracture surfaces signified expanded fracture area and complex fracture paths, which were believed to dissipate more energy and thus improve toughness.32 The remarkable strengthening and toughening effect of VAD was related to the non-phase separation morphology and the perfect combination of rigid and flexible groups. Firstly, the formation of block copolymers of VAD and EP can improve the energy dissipation in the failure process, and the introduction of plentiful aromatic rings, which can generate π–π interaction with the matrix further improved the interaction between VAD and the matrix.33 Secondly, the flexible chain of VAD increased the movement ability between molecular segments, increasing the likelihood of homogeneous shear deformation and dissipating more fracture energy via intermolecular motion. Therefore, VAD played a toughening effect in the mode of energy absorption described as “in situ toughening”.34
![]() | ||
Fig. 9 SEM images of the fracture surfaces of the samples: (a) pure EP, (b) EP/VAD-2.5 and (c) EP/VAD-5 at different magnification. |
Samples | EP (g) | MACA (g) | VAD (g) | P (wt%) |
---|---|---|---|---|
EP | 100 | 26.50 | 0 | 0 |
EP/VAD-2.5 | 100 | 25.85 | 3.23 | 0.15 |
EP/VAD-5 | 100 | 25.21 | 6.59 | 0.30 |
EP/VAD-7.5 | 100 | 24.55 | 10.09 | 0.45 |
1H NMR and 31P NMR spectra were acquired, using DMSO-d6 as the solvent, using an NMR spectrometer, Bruker AV400, Switzerland.
Transmission electron microscopy (TEM) of samples was carried out using a transmission electron microscope (JEM-1400Plus, Japan) at 120 kV. Scanning electron microscopy (SEM) was utilized to acquire the fracture morphology of the tensile specimen and the failed surface micro-morphology caused by the impact with LOX using a scanning electron microscope (JSM-7500F, Japan) under the activation voltage of 20 kV. An energy-dispersive X-ray spectrometer (EDX) was fitted for the P element investigation in the surface scanning model.
Differential scanning calorimetry (DSC) measurement (heating rate = 10 °C min−1, 30–250 °C, N2 atmosphere) was performed using a PerkinElmer DSC 4000 instrument, USA.
Thermogravimetric analysis (TGA) was performed using a NETZSCH STA449F3 instrument, Germany (heating rate = 10 °C min−1, 30–800 °C, N2 atmosphere).
The LOX compatibility was tested under 98 J impact energy using the Army Ballistic Missile Agency (ABMA)-type impact tester according to ASTM D2512-95.35 A specimen of the test material, whose dimension was 20 mm in diameter and 2 mm thick, was placed in a specimen cup, pre-cooled with LOX for 10 min. The 10 kg hammer was released from a height of 1 meter, which transmitted 98 J energy to the test sample. According to whether any LOX-impact sensitive response (burning, explosion, flash or charring) occurred during the test and the number of sensitive phenomena was used to characterize the impact reaction sensitivity (IRS). A material subjected to 20 impact tests with no response would be considered compatible with LOX.
The limited oxygen index (LOI) values were determined using an ASTM D2863-compliant JF-3 oxygen index meter (Jiangning, China), and all samples were 100 × 6.5 × 3 mm3.
X-ray photoelectron spectroscopy (XPS) technique was used to measure the surface element compositions at 100–1400 eV of the cured EPs before and after the LOX impact using X-ray photoelectron spectroscopy, ESCALAB 250Xi instrument, USA.
Tensile and flexural properties were conducted according to GB/T 2567–2008 using a universal testing machine (Instron 5967, USA) at a 2 mm min−1 loading rate; the impact strength was executed on a pendulum impact tester (XJJD-50, China). Critical stress intensity factors (KIC) were assessed using a universal testing machine (SHT4106, USA) in the three-point bending test at a 1 mm min−1 loading rate according to ASTM D5045-14. The sample dimension was 35.2 × 8 × 4 mm3 and the length of the prefabricated crack was 3.6 mm. Each of the above tests went through five parallel tests, and the obtained value was the average of the five tests.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/d1ra09049f |
This journal is © The Royal Society of Chemistry 2022 |