A New DOPO-Eugenol Adduct as an Effective Flame Retardant for Epoxy Thermosets with Improved Mechanical Properties

The development of efficient green flame retardants is an important way to realize more sustainable epoxy thermosets and downstream materials. In this work, a monoepoxide is synthesized through O-glycidylation of eugenol, and then reacted with DOPO (9,10-dihydro-9-oxa-10-phosphophenanthrene-10-oxide) to obtain a new bio-based flame retardant, DOPO-GE. DOPO-GE is blended with a bisphenol A epoxy prepolymer exhibiting good compatibility and DDS (4,4′-diaminodiphenylsulfone) is used as the curing agent to afford epoxy thermosets. Although DOPO-GE leads to the reduced glass transition temperature of the thermosets, the storage modulus increases considerably. The DOPO-GE-modified thermosets exhibit the high thermal stability with the onset thermal decomposition temperature in nitrogen and air exceeding 300°C. When the phosphorus content in the thermoset is 1.0%, the residual yield of the thermosets at 750°C in nitrogen increases from 13.9% to 30.6%, due to the increased charring ability. More interestingly, when the phosphorus content is only 0.5%, the limiting oxygen index is as high as 30.3% with UL94 V0 achieved. Cone calorimeter results reveals the significantly decreased heat release rate, total heat release, mass loss and total smoke production. Furthermore, DOPO-GE can notably improve the flexural strength, flexural modulus and fracture toughness, whereas the shear and impact strength are reduced to varied extents. In short, DOPO-GE can be obtained via a facile way, and shows the good flame-retardant effect on the epoxy thermosets with an application potential.


Introduction
Epoxy resins are widely used in many fields mainly because of their excellent mechanical properties, insulation properties and good processability. Bisphenol A-based epoxy resin (DGEBA) is the most commonly used epoxy prepolymer, accounting for more than 80% market share of the whole epoxy resin market. However, DGEBA is flammable and easy to burn in air. Therefore, improving the flame retardancy of DGEBA is a very important issue for a number of applications, and thus becomes a hot research topic in academia and industry. Many methods have been used to improve the flame retardancy of DGEBA. For example, brominated epoxy resins have high intrinsic flame retardancy, but they produce highly toxic gases in combustion. Therefore, halogen-free flame retardants for epoxy is of particular importance. For example, aluminum hydroxide, magnesium hydroxide and other inorganic fillers are added to the epoxy systems to improve flame retardancy, but such inorganic flame retardants usually need a high loading, resulting in problems regarding processing and compatibility. Therefore, developing organic flame retardants with better compatibility and higher efficiency provides an alternative solution to the above problems. In particular, organic flame retardants based on DOPO (9,10-dihydro-9-oxa-10-phosphophenanthrene-10-oxide) have the advantages of high flame retardancy, low volatilization, low moisture absorption, good transparency, good compatibility, and reasonable costs, so that DOPO functions as the very useful building block to develop many new and efficient flame retardants for polymeric materials including epoxy thermosets [1][2][3][4][5][6][7][8].
In order to reduce the impact on environment and relieve resource shortage, the development of renewable epoxy resin and related flame retardants has attracted more and more attention. In recent years, many studies on the synthesis of epoxy thermosets and related flame retardants by replacing bisphenol A with bio-based phenols have been reported extensively, such as cardanol [9][10][11][12][13], vanillin [14][15][16], magnolol [17,18], ellagic acid [19], syringaldehyde [20], and eugenol, to name a few. Among them, eugenol is of particular interest. Eugenol is a natural substituted phenolic compound. Its molecules contain such active groups as allyl and phenolic hydroxyl groups. Eugenol can be chemically functionalized to obtain a variety of bio-based thermoplastics and thermosets, additives, and functional polymeric materials . At the same time, a number of publications address eugenol-based flameretardant epoxy systems. For example, the phenolic group of eugenol is used to react with different linkers such as phosphorus containing organic compounds to produce di-, tri-or multi-allyl compounds, and then the corresponding di-, tri-or multi-functional epoxy monomers or oligomers are obtained by epoxidation of the allyl groups [63][64][65][66]. Certainly, DOPO is an efficient epoxy flame retardant, but to our knowledge, DOPO-modified eugenol is still not applied as the flame retardant for epoxy thermosets.
In this paper, eugenol is undergone O-glycidylation to introduce a highly reactive terminal epoxy group, and then reacted with phosphorus hydrogen bond of DOPO to obtain a DOPO-modified eugenol-based flame retardant (DOPO-GE). DOPO-GE is used to modify commercially available DGEBA, and 4,4′diaminodiphenylsulfone (DDS) is used as the curing agent to obtain the thermosets. We systematically study the dynamic mechanical properties, thermal decomposition behavior, flame retardancy, mechanical properties and fracture behavior of the thermosets. It is found that DOPO-GE is very effective to improve the flame retardancy and can increase the rigidity of the obtained epoxy materials simultaneously.

Reagents and Raw Materials
Eugenol was purchased from a commercial source, and decolorized by vacuum distillation before use. Fresh-distilled eugenol is a colorless liquid with a purity of +99% (GC). Epichlorohydrin, sodium hydroxide, benzyltriethylammonium chloride and methanol were purchased from Sinopharm. 4,4′-Diaminediaminodiphenyl sulfone (99%) was obtained from Energy Chemical. Commodity bisphenol A epoxy resin (E54 TM ) with epoxy value of 0.556 mol/100 g was used in this study. Unless there were other special specifications, all the chemicals and materials were used as received.

Synthesis of Flame Retardants
As shown in Scheme 1, add eugenol (100 g, 0.61 mol), 2.53 g of benzyltriethylammonium chloride and epichlorohydrin (500 ml) into a 1000 ml flask in a water bath. After reacting at 80°C for 6 hours, the flask was rotary evaporated to remove unreacted epichlorohydrin up to <10 mbar for 30 min. The remainder was dissolved into 500 ml of toluene, and then 40% sodium hydroxide solution (24.4 g of NaOH) was added dropwise at room temperature with stirring for 6 h. The upper organic layer was separated, extracted with deionized water several times until neutrality, and dried over anhydrous sodium sulfate overnight. After removal of the solvent in a vacuum oven at 80°C overnight, the obtained crude product was dissolved in methanol and chilled to yield white crystals. The collected crystals was dried in a vacuum to obtain glycidyl ether of eugenol (GE) in a~70% yield. DOPO (62.30 g, 0.28 mol) and glycidyl ether of eugenol (62.40 g, 0.28 mol) were charged into a 500 ml flask and heated in an oil bath (170°C) with stirring to obtain a clear solution. Then triphenylphosphine (0.31 g) was added and reacted for 28 h to obtain a viscous liquid without further separation as the biobased flame retardant (DOPO-GE) in a quantitative yield.

Preparation of Cured Thermosets
To a preheated (135°C) mold sprayed with releasing agent, a quantitative mixture sample of E54, DDS, and DOPO-GE were filled and degassed under reduced pressure. The following curing procedures were carried out: 180°C for 2 h, 200°C for 4 h, and 220°C for 4 h, and then cooled to room temperature. Disassemble the mold and machine the cured thermosets into desired dimensions for further testing. The typical formulations of the epoxy thermosets are listed in Table 1.

Characterizations
Infrared spectroscopy (FT-IR). A spectrophotometer (PerkinElmer Spectrum Two UATR) was used to register IR spectra. Potassium bromide was used as the supporter and wavenumber range was 4000-500 cm −1 .
Nuclear magnetic resonance (NMR). A nuclear magnetic resonance spectrometer (400 MHz, JEOL) was used to acquire the 1 H NMR spectra of the samples with deuterated chloroform (CDCl 3 ) as the solvent and tetramethylsilane as the internal standard, and the number of scans was 8 times.
Thermogravimetric analysis. A thermal analysis system instrument (Q600, TA Instruments) was used to analyze the thermal decomposition of the cured epoxy samples. An appropriate amount (4-5 mg) of the cured epoxy thermoset was heated from room temperature at a heating rate of 20 K/min to 800°C in an air and a nitrogen atmospheres, respectively.
Dynamic mechanical thermal analysis (DMA). A DMA analyzer (Q800, TA instruments) was used to analyze the viscoelasticity of the cured epoxy thermosets. The sample (60 mm × 10 mm × 2.6 mm) was fixed Scheme 1: Synthesis of glycidyl ether of eugenol and its DOPO adduct on the dual-cantilever beam, the amplitude was 25 μm, the frequency was 1 Hz, and the heating rate was 3 K/min from room temperature to 300°C.
Shear strength. A universal testing machine (RGM-2010, Regal Instrument, Ltd., China) was used to measure the shear strength according to ISO 527-2012 with the tensile speed of 10 mm/min. The epoxy adhesive was coated on a steel plate with bonding surface of 12.5 mm × 25 mm, and cured at 180°C for 6 h.
Flexural properties. Testing of flexural properties of the cured epoxy was conducted on the universal testing machine. According to the national standard GB/T 9341-2008, the polished specimens were measured at room temperature with the crosshead speed of 5 mm/min. Impact strength. An impact tester (XJJD-5, Chengde Jinjian Testing Instrument, China) was used to determine impact strength of V-notched specimens according to GB/T 2571-1995. The impact speed was 2.9 m/s and the impact energy (2 J).
Fracture toughness and fracture energy. The universal testing machine was used to measure the fracture toughness and fracture energy according to ASTM D5045. The processed samples (length 60 ± 0.02 mm, width 10 ± 0.02 mm, thickness 5 ± 0.02 mm, and notch 4 ± 0.02 mm) were tested at the crosshead speed of 10 mm/min in bending mode.
Limiting oxygen index (LOI). An oxygen index tester (JF-5, Fujian Survey Instrument and Equipment, China) was used to determine the LOI values in terms of ISO 4589-1996. UL 94 vertical burning test was carried out on the cured epoxy samples. A cone calorimeter (Fire Testing Technology) test was used to study flame retardancy the samples (100 mm × 100 mm × 4 mm) at a fixed heat flux of 35 kW/m 2 according to ISO 5660.
Scanning electron microscope (SEM). A SEM equipment (SU3500, Hitachi, Japan) was used to check the surface morphology of the residue of the samples after burning in the air. Gold was sprayed on the surface of the samples and the accelerating voltage applied was 5 KV.

Molecular Characterization
As shown in Fig. 1A, DOPO has the infrared absorption at~2400 cm −1 due to the stretching vibration of phosphorus hydrogen (P-H) bonds, whereas DOPO-GE shows no absorption associated with P-H bond. Meanwhile, for DOPO-GE there is a strong hydroxyl stretching vibration between 3500-3000 cm −1 , which is due to the ring-opening reaction between the epoxy group and the P-H bond [67]. Moreover, DOPO-GE displays a strong stretching vibration owing to the carbon-carbon double bond at~1632 cm −1 , indicating that the allyl double bond of GE has not changed. In addition, from 1 H NMR spectra (Fig. 2B) the signals of the allyl double on the product are observed at about 4.9 and 5.8 ppm, and the resonance of P-H bond at about 8.9 ppm in DOPO disappears, which further confirms that the P-H bond of DOPO has reacted with epoxy group of GE via a ring-opening process.  Fig. 2 shows the relationship between the dynamic mechanical properties (storage modulus E′ and loss factor Tan δ) and temperature of the thermosets, E54/DDS, E54/DDS/1.0%P and E54/DDS/2.0%P. In a certain temperature range (below~110°C), E′ of E54/DDS/1.0%P and E54/DDS/2.0%P is higher than that of pristine E54/DDS, and increases with the DOPO-GE contents. For example, at 25°C, E′ of E54/ DDS/2.0%P is the highest up to 3421 MPa, while the E′ of DGEBA/DDS is the lowest (2339 MPa). At 100°C, E′ of E54/DDS/1.0%P and E54/DDS/2.0%P systems exceeds 2.1 GPa, while the E′ of E54/DDS system does not exceed 1.6 GPa. Therefore, the addition of DOPO-GE can increase the rigidity of the cured thermosets within a relatively lower temperature range. The reason lies in that DOPO-GE provides condensed rigid aromatic segments with a stronger intermolecular interaction, and the allyl groups could react into the thermosetting network, which makes the motion of the molecular segments more difficult. Fig. 2 shows that with the further increase of temperature, E′ of thermoset decreases greatly in a certain temperature range, and the dissipation factor (Tan δ) and loss modulus (E′′) increase dramatically reaching their maximum, and then decreases rapidly approaching zero. This observation mean that the materials undergo the glass transition. The temperature for Tan δ max is taken as the glass transition temperature (T g ). It can be found that with the increase of DOPO-GE fraction, Tan δ max of decreases gradually, while the glass-transition temperature range widens. Correspondingly, as shown in Table 2, T g decreases from 238.6°C for E54/DDS to 143.8°C for the thermoset with 2.0%P content, indicating that DOPO-GE has strong plasticizing effect to tailor the glass transition temperature of the resultant thermosets. Interestingly, E54/DDS/1.0%P and E54/DDS/2.0%P show a higher rubbery modulus, likely due to allyl's polymerization which produces additional crosslinks [18].

Thermal Decomposition Behaviors
The effect of DOPO-GE on the thermal decomposition behavior of the cured epoxy resin in nitrogen and air is studied, see Fig. 3. The characteristic thermal decomposition temperatures, such as the onset, 5%, the thermal decomposition temperatures corresponding to the maximum decomposition rate (T onset , T d5 and T dmax ), and the residual fraction at 750°C are listed in Table 3. With the increased fraction of DOPO-GE, T onset decreases likely due partially to the dissociation of the methyl functions inherited from DOPO-GE [67]. Nevertheless, all the systems have high thermal stability, no apparent thermal degradation occurs below 250°C, and T d5 and T onset exceed 300°C. These results show that the thermal stability of the material can fully meet the requirements as a hard plastic, because the decomposition temperature is much higher than T g . Specifically, in air, with the increase of DOPO-GE loading, the thermal decomposition temperatures decrease gradually, whereas the residual fraction shows a trend of first increasing and then decreasing. Accordingly, when the phosphorus content is 1.0%, the thermoset reaches its maximum residual yield, so DOPO-GE has an optimal content. The DOPO units from DOPO-GE can catalyze the dehydration of the epoxy. At the same time, organic phosphorus will transform into phosphoric acids after pyrolyzation, which can significantly catalyze the carbonization of the polymer matrix [68]. In air, the respective characteristic thermal decomposition temperatures of the systems, in general, decrease as compared with the case of N 2 , and the effect of phosphorus content is also similar to that of the system in N 2 during the first-step of decomposition. However, the thermal decomposition in the air shows two distinct steps. After the first step, the epoxy is carbonized, and the carbon formed in the second step will burn in the air and produce soot.

Flammability Analysis
The flame retardancy of the epoxy materials is characterized by limiting oxygen index (LOI) and from UL94 vertical burning tests. See Table 3 for the detailed results. When the epoxy contains only 0.5 wt% phosphorus (E54/DDS/1.0%P), its LOI increases from 22.6% for the unmodified (E54/DDS) to 30.3%. At the same time, the corresponding UL94 classification is greatly enhanced from no level to V0 grade. It can be concluded that DOPO-GE is an effective flame retardant for epoxy under the condition of relatively low loadings. When the phosphorus content in the system increased to 1.0 wt%, the flame retardancy of the materials is further improved, and the LOI is as high as 32.5%. However, as the loading of DOPO-GE continues to increase, the LOI decreases. This finding suggests that the phosphorus content has an optimal value, about 1.0%. Because too high DOPO-GE loading may weaken the carbonization ability of the thermosets and reduce the strength of the carbon structure formed during combustion.
By using a cone calorimeter to monitor combustion of the epoxy, the characteristic parameters of the thermosets containing the flame retardants with the 0.5%P content are studied in a more quantitative and comparable way. Fig. 4 shows how the HRR (heat release rate), THR (Total heat release), mass loss and TSP (Total smoke production) changes with the heating time, and Table 4 lists the main results from this test. In Fig. 4A and Table 4, pristine E54/DDS displays a slighter higher ignition time than E54/DDS/0.5%P does, due to the easy thermal decomposition of the flame retardant. However, the peak HRR value of E54/DDS/0.5%P is lower than that of E54/DDS (659 ± 45 vs. 831 ± 50 kW m −2 ), with a 15.9% difference. The larger the HRR, the higher the fire intensity and the greater the danger of the material when catching fire. Thus, E54/DDS/0.5%P has the better flame retardancy than E54/DDS does.
In Fig. 4B, the THR of the thermoset increases over time, and eventually stabilizes indicting extinguishing. After 254 s, the THR value of the E54/DDS/0.5%P system is lower than that of E54/DDS, and finally reduces by 25.8%, implicating a reduction in the production of combustible volatiles. In Fig. 4C, E54/DDS/0.5%P shows a lower mass loss than DGEBA/DDS (65.99% vs. 86.75%) does, decreased by 24.0%. Because DOPO moieties in epoxy thermoset could catalyze the charring with the reduced gaseous combustible formation. The smoke generation is compared in Fig. 4D and Table 4. E54/DDS/0.5%P shows a decreased total smoke production (TSP) than E54/DDS does after~290 s. As the ignition time increases, TSP increases rapidly, and finally the increase becomes sluggish and TSP tends to level off. Moreover, E54/DDS/0.5%P shows the much lower total smoke release (TSR) and total smoke production (TSP) values than E54/DDS does with a 22.2% reduction, indicating the decreased risk of fire hazard by suppressing smoke generation. The mechanisms of the enhanced flame retardancy and smoke depression are associated with increased charring ability and formation of protective char layers during combustion to form a tortuous pathway which slows down the mass and heat transfer and the free radical inhibition being in action in a gaseous phase [68][69][70][71][72].  Fig. 5 compares the effect of DOPO-GE on the epoxy thermosets with the different phosphorus content of 0, 1.0% and 1.5%, by highlighting the morphology of the residual obtained after combustion in air. Without addition of DOPO-GE, E54/DDS presents many large holes, due to the melting of the sample during the combustion process. In this case, once the sample is ignited, the generated heat will cause  rapid polymer chain scission leading to the melting and hole formation and produce a large amount of flammable gases. In this way, the diffusion of flammable gases and air and the heat transfer are greatly accelerated, which will promote the further decomposition and burning of the remaining epoxy matrix. After combustion in air, E54/DDS/1.0%P and E54/DDS/1.5%P show intumescent sponge-like morphology. The higher phosphorus content of the systems leads to the denser sponge morphology. This finding is likely due to the increased charring ability of the system to form porous carbon during the combustion process. The increased charring ability of the modified epoxy thermoset is due to the catalytic effect of phosphoric acid species formed during the pyrolysis. The formation of intumescent carbon structure will retard the diffusion of the flammables and heat transfer to underneath resin matrix, which will greatly slow down the combustion and even lead to self-extinguishing. These findings provide the further evidence for the better flame retardancy of the DOPO-GE-modified epoxy systems [8,[73][74][75].

Mechanical Performances of Thermosets
The mechanical properties of the thermosets are evaluated, and the effects of DOPO-GE on shear strength, impact strength, flexural modulus and fracture toughness and energy are investigated. The results in Table 5 show that with the increase of DOPO-GE loading, the flexural strength of the material increases. Especially, when the phosphorus content is 0.5 wt%, the flexural strength and modulus of the material increases from 183 to 281 MPa and from 2.34 to 3.98 GPa, respectively. Flexural properties of the material are improved, which is due to the improvement of the rigidity of the epoxy network. This is consistent with the results from the DMA results (Fig. 2). The impact strength decreases with an increase in the phosphorus content, which indicates that the plastic deformation of the material decreases under a high-speed impact stress. This may be related to the strong cohesive energy density of DOPO-GE in the epoxy system, which leads to an increase in the flexural properties, but makes the viscosity relaxation of the polymer chains more difficult. The decreased shear strengths are more likely related to the properties of bonding surface such as wettability and surface energy. Also, the effects of DOPO-GE are investigated in terms of the fracture toughness (K IC ) and fracture energy (G IC ) of the thermosets. K IC increases systematically from 2.4 to 4.9 MPa · m 1/2 , as the phosphorus content increases from 0 to 2.0%. Meanwhile, G IC arrives at the maximum of 5.7 MPa·m 1/2 at 1.0% phosphorus content. This may be related to the increased the intermolecular interaction of the material to inhibit propagation of the cracks under a bending stress. Taken together, under the condition of the good flame-retardant performance, the phosphorus content of 0.5%-1.0% is a better choice to achieve well-balanced mechanical properties of the epoxy thermosets.

Conclusions
By reacting glycidyl ether of eugenol with DOPO, a new bio-based halogen-free phosphorus flame retardant (DOPO-GE) was readily synthesized. Its molecular structure was characterized by FTIR and 1 H NMR. DOPO-GE was used to modify a bisphenol A epoxy prepolymer (E54) with DDS as the curing agent. DOPO-GE significantly reduced the glass transition temperature of the resultant thermosets, but increased the storage modulus. The cured epoxy material had high thermal stability. The onset thermal decomposition temperature in nitrogen and air exceeded 300°C. When the phosphorus content was 1.0 wt%, the residual content at 750°C in nitrogen increased from 13.9% for the unmodified epoxy to 30.6%. Therefore, the charring ability of the epoxy was greatly enhanced. When the phosphorus content was as low as 0.5 wt%, the LOI value could reach 30.3%, and the thermosets could pass the UL94 test with V0 level arrived. When the content of phosphorous content was increased to 1.0 wt%, LOI could further increase to 32.5%. Cone calorimeter results showed that the 0.5 wt% phosphorus incorporation in the epoxy led to the significantly decreased HRR (heat release rate), THR (total heat release) and TSP (Total smoke production), by 15.9%, 25.5% and 22.2%, respectively. The results of mechanical properties showed that the addition of DOPO-GE could significantly improve the flexural strength, modulus, fracture toughness, but could decrease the impact and shear strengths. In short, DOPO-GE is an effective halogen-free flame retardant derivable from renewable eugenol. When its loading is low (0.5-1.0 wt%, phosphorus based), DOPO-GE can effectively improve flame retardancy of the epoxy thermosets, and can enhance the rigidity of the epoxy network at the same time.

Conflicts of Interest:
The authors declare that they have no conflicts of interest to report regarding the present study.